All Classes and Interfaces

Class
Description
Raise a exception to abort the process when called.
 
Optional attributes for Abort
Computes the absolute value of a tensor.
 
Abstract base class for Activations
Base class for Constraints.
Helper base class for custom gradient adapters INTERNAL USE ONLY
Base class for Regularizers
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Returns the element-wise sum of a list of tensors.
 
Applies a gradient to a given accumulator.
 
Returns the number of gradients aggregated in the given accumulators.
 
Updates the accumulator with a new value for global_step.
 
Extracts the average gradient in the given ConditionalAccumulator.
 
Metric that calculates how often predictions equals labels.
Computes acos of x element-wise.
 
Computes inverse hyperbolic cosine of x element-wise.
 
Interface for Activations
The Enumerations for creating Activations based an activation name, with either an empty constructor or a constructor that takes a Map object that contains the Activation's state.
Optimizer that implements the Adadelta algorithm.
Optimizer that implements the Adagrad algorithm.
Optimizer that implements the Adagrad Dual-Averaging algorithm.
Optimizer that implements the Adam algorithm.
Optimizer that implements the Adamax algorithm.
Returns x + y element-wise.
 
Add an N-minibatch SparseTensor to a SparseTensorsMap, return N handles.
 
Optional attributes for AddManySparseToTensorsMap
Add all input tensors element wise.
 
Add a SparseTensor to a SparseTensorsMap return its handle.
 
Optional attributes for AddSparseToTensorsMap
Adjust the contrast of one or more images.
 
Adjust the hue of one or more images.
 
Adjust the saturation of one or more images.
 
Computes the "logical and" of elements across dimensions of a tensor.
 
Optional attributes for All
Generates labels for candidate sampling with a learned unigram distribution.
 
Optional attributes for AllCandidateSampler
Protobuf type tensorflow.AllocationDescription
Protobuf type tensorflow.AllocationDescription
 
 
An allocation/de-allocation operation performed by the allocator.
An allocation/de-allocation operation performed by the allocator.
 
Protobuf type tensorflow.AllocatorMemoryUsed
Protobuf type tensorflow.AllocatorMemoryUsed
 
An Op to exchange data across TPU replicas.
 
Returns the argument of a complex number.
 
Creates a uninitialized anonymous hash table.
 
A container for an iterator resource.
 
The AnonymousMemoryCache operation
 
A container for a multi device iterator resource.
 
Creates an empty anonymous mutable hash table that uses tensors as the backing store.
 
Optional attributes for AnonymousMutableDenseHashTable
Creates an empty anonymous mutable hash table.
 
Creates an empty anonymous mutable hash table of vector values.
 
Optional attributes for AnonymousMutableHashTableOfTensors
The AnonymousRandomSeedGenerator operation
 
The AnonymousSeedGenerator operation
 
Computes the "logical or" of elements across dimensions of a tensor.
 
Optional attributes for Any
Used to specify and override the default API & behavior in the generated code for client languages, from what you would get from the OpDef alone.
Protobuf type tensorflow.ApiDef.Arg
Protobuf type tensorflow.ApiDef.Arg
 
Description of the graph-construction-time configuration of this Op.
Description of the graph-construction-time configuration of this Op.
 
Used to specify and override the default API & behavior in the generated code for client languages, from what you would get from the OpDef alone.
If you specify any endpoint, this will replace all of the inherited endpoints.
If you specify any endpoint, this will replace all of the inherited endpoints.
 
Protobuf enum tensorflow.ApiDef.Visibility
 
 
Protobuf type tensorflow.ApiDefs
Protobuf type tensorflow.ApiDefs
 
Update '*var' according to the adadelta scheme. accum = rho() * accum + (1 - rho()) * grad.square(); update = (update_accum + epsilon).sqrt() * (accum + epsilon()).rsqrt() * grad; update_accum = rho() * update_accum + (1 - rho()) * update.square(); var -= update;
 
Optional attributes for ApplyAdadelta
Update '*var' according to the adagrad scheme. accum += grad * grad var -= lr * grad * (1 / sqrt(accum))
 
Optional attributes for ApplyAdagrad
Update '*var' according to the proximal adagrad scheme.
 
Optional attributes for ApplyAdagradDa
Update '*var' according to the adagrad scheme. accum += grad * grad var -= lr * grad * (1 / sqrt(accum))
 
Optional attributes for ApplyAdagradV2
Update '*var' according to the Adam algorithm.
 
Optional attributes for ApplyAdam
Update '*var' according to the AdaMax algorithm. m_t <- beta1 * m_{t-1} + (1 - beta1) * g v_t <- max(beta2 * v_{t-1}, abs(g)) variable <- variable - learning_rate / (1 - beta1^t) * m_t / (v_t + epsilon)
 
Optional attributes for ApplyAdaMax
Update '*var' according to the AddSign update. m_t <- beta1 * m_{t-1} + (1 - beta1) * g update <- (alpha + sign_decay * sign(g) *sign(m)) * g variable <- variable - lr_t * update
 
Optional attributes for ApplyAddSign
Update '*var' according to the centered RMSProp algorithm.
 
Optional attributes for ApplyCenteredRmsProp
Update '*var' according to the Ftrl-proximal scheme. grad_with_shrinkage = grad + 2 * l2_shrinkage * var accum_new = accum + grad * grad linear += grad_with_shrinkage - (accum_new^(-lr_power) - accum^(-lr_power)) / lr * var quadratic = 1.0 / (accum_new^(lr_power) * lr) + 2 * l2 var = (sign(linear) * l1 - linear) / quadratic if |linear| > l1 else 0.0 accum = accum_new
 
Optional attributes for ApplyFtrl
Update '*var' by subtracting 'alpha' * 'delta' from it.
 
Optional attributes for ApplyGradientDescent
Update '*var' according to the momentum scheme.
 
Optional attributes for ApplyMomentum
Update '*var' according to the AddSign update. m_t <- beta1 * m_{t-1} + (1 - beta1) * g update <- exp(logbase * sign_decay * sign(g) * sign(m_t)) * g variable <- variable - lr_t * update
 
Optional attributes for ApplyPowerSign
Update '*var' and '*accum' according to FOBOS with Adagrad learning rate. accum += grad * grad prox_v = var - lr * grad * (1 / sqrt(accum)) var = sign(prox_v)/(1+lrl2) * max{|prox_v|-lrl1,0}
 
Optional attributes for ApplyProximalAdagrad
Update '*var' as FOBOS algorithm with fixed learning rate. prox_v = var - alpha * delta var = sign(prox_v)/(1+alphal2) * max{|prox_v|-alphal1,0}
 
Optional attributes for ApplyProximalGradientDescent
Update '*var' according to the RMSProp algorithm.
 
Optional attributes for ApplyRmsProp
Returns the truth value of abs(x-y) < tolerance element-wise.
 
Optional attributes for ApproximateEqual
Returns min/max k values and their indices of the input operand in an approximate manner.
 
Optional attributes for ApproxTopK
Returns the index with the largest value across dimensions of a tensor.
 
Returns the index with the smallest value across dimensions of a tensor.
 
Computes the trignometric inverse sine of x element-wise.
 
Computes inverse hyperbolic sine of x element-wise.
 
The AssertCardinalityDataset operation
 
A transformation that asserts which transformations happen next.
The ExperimentalAssertNextDataset operation
 
 
A transformation that asserts which transformations happened previously.
 
Asserts that the given condition is true.
 
Optional attributes for AssertThat
An asset file def for a single file or a set of sharded files with the same name.
An asset file def for a single file or a set of sharded files with the same name.
 
Update 'ref' by assigning 'value' to it.
 
Optional attributes for Assign
Update 'ref' by adding 'value' to it.
 
Optional attributes for AssignAdd
Adds a value to the current value of a variable.
 
Update 'ref' by subtracting 'value' from it.
 
Optional attributes for AssignSub
Subtracts a value from the current value of a variable.
 
Concats input tensor across all dimensions.
 
Optional attributes for AssignVariableConcatND
Assigns a new value to a variable.
 
Optional attributes for AssignVariableOp
Converts each entry in the given tensor to strings.
 
Optional attributes for AsString
Computes the trignometric inverse tangent of x element-wise.
 
Computes arctangent of y/x element-wise, respecting signs of the arguments.
 
Computes inverse hyperbolic tangent of x element-wise.
 
Metadata of an op's attribute.
Protocol buffer representing the value for an attr used to configure an Op.
Protocol buffer representing the value for an attr used to configure an Op.
LINT.IfChange
LINT.IfChange
 
 
 
 
Metric that computes the approximate AUC (Area under the curve) via a Riemann sum.
Specifies the type of the curve to be computed, AUCCurve.ROC for a Receiver Operator Characteristic curve [default] or AUCCurve.PR for a Precision-Recall-curve.
Specifies the Riemann summation method used.
An API for building audio operations as Ops
Produces a visualization of audio data over time.
 
Optional attributes for AudioSpectrogram
Outputs a Summary protocol buffer with audio.
 
Optional attributes for AudioSummary
Protobuf type tensorflow.AutoParallelOptions
Protobuf type tensorflow.AutoParallelOptions
 
Creates a dataset that shards the input dataset.
Creates a dataset that shards the input dataset.
 
 
Optional attributes for AutoShardDataset
Optional attributes for AutoShardDataset
Matches DeviceAttributes
Matches DeviceAttributes
 
Performs average pooling on the input.
 
Optional attributes for AvgPool
Performs 3D average pooling on the input.
 
Optional attributes for AvgPool3d
Computes gradients of average pooling function.
 
Optional attributes for AvgPool3dGrad
Computes gradients of the average pooling function.
 
Optional attributes for AvgPoolGrad
Axes Operations
The BandedTriangularSolve operation
 
Optional attributes for BandedTriangularSolve
Copy a tensor setting everything outside a central band in each innermost matrix to zero.
 
Defines a barrier that persists across different graph executions.
 
Optional attributes for Barrier
Closes the given barrier.
 
Optional attributes for BarrierClose
Computes the number of incomplete elements in the given barrier.
 
For each key, assigns the respective value to the specified component.
 
Computes the number of complete elements in the given barrier.
 
Takes the given number of completed elements from a barrier.
 
Optional attributes for BarrierTakeMany
Abstract base class for all Initializers
Base class for Metrics
Batches all input tensors nondeterministically.
 
Optional attributes for Batch
The BatchCholesky operation
 
The BatchCholeskyGrad operation
 
Creates a dataset that batches batch_size elements from input_dataset.
 
Optional attributes for BatchDataset
The BatchFFT operation
 
The BatchFFT2D operation
 
The BatchFFT3D operation
 
Batches all the inputs tensors to the computation done by the function.
 
Optional attributes for BatchFunction
The BatchIFFT operation
 
The BatchIFFT2D operation
 
The BatchIFFT3D operation
 
Multiplies slices of two tensors in batches.
 
Optional attributes for BatchMatMul
The BatchMatrixBandPart operation
 
The BatchMatrixDeterminant operation
 
The BatchMatrixDiag operation
 
The BatchMatrixDiagPart operation
 
The BatchMatrixInverse operation DEPRECATED: This operation is deprecated and will be removed in a future version.
 
Optional attributes for BatchMatrixInverse
The BatchMatrixSetDiag operation
 
The BatchMatrixSolve operation
 
Optional attributes for BatchMatrixSolve
The BatchMatrixSolveLs operation
 
Optional attributes for BatchMatrixSolveLs
The BatchMatrixTriangularSolve operation
 
Optional attributes for BatchMatrixTriangularSolve
Batch normalization.
 
Gradients for batch normalization.
 
The BatchSelfAdjointEigV2 operation
 
Optional attributes for BatchSelfAdjointEig
The BatchSvd operation
 
Optional attributes for BatchSvd
BatchToSpace for 4-D tensors of type T.
 
BatchToSpace for N-D tensors of type T.
 
Protobuf type tensorflow.BenchmarkEntries
Protobuf type tensorflow.BenchmarkEntries
 
Each unit test or benchmark in a test or benchmark run provides some set of information.
Each unit test or benchmark in a test or benchmark run provides some set of information.
 
The BesselI0 operation
 
The BesselI0e operation
 
The BesselI1 operation
 
The BesselI1e operation
 
The BesselJ0 operation
 
The BesselJ1 operation
 
The BesselK0 operation
 
The BesselK0e operation
 
The BesselK1 operation
 
The BesselK1e operation
 
The BesselY0 operation
 
The BesselY1 operation
 
Compute the regularized incomplete beta integral \(I_x(a, b)\).
 
 
 
Protobuf type tensorflow.BinSummary
Protobuf type tensorflow.BinSummary
 
Some of the data from AllocatorStats
Some of the data from AllocatorStats
 
Protobuf type tensorflow.MemChunk
Protobuf type tensorflow.MemChunk
 
Protobuf type tensorflow.MemoryDump
Protobuf type tensorflow.MemoryDump
 
Protobuf type tensorflow.SnapShot
Protobuf type tensorflow.SnapShot
 
Adds bias to value.
 
Optional attributes for BiasAdd
The backward operation for "BiasAdd" on the "bias" tensor.
 
Optional attributes for BiasAddGrad
Metric that calculates how often predictions matches binary labels.
Computes the cross-entropy loss between true labels and predicted labels.
A Metric that computes the binary cross-entropy loss between true labels and predicted labels.
Counts the number of occurrences of each value in an integer array.
 
Bitcasts a tensor from one type to another without copying data.
 
Elementwise computes the bitwise AND of x and y.
 
An API for building bitwise operations as Ops
Elementwise computes the bitwise OR of x and y.
 
Elementwise computes the bitwise XOR of x and y.
 
Computes the LSTM cell forward propagation for all the time steps.
 
Optional attributes for BlockLSTM
Computes the LSTM cell backward propagation for the entire time sequence.
 
 
Optional attributes for BooleanMask
 
Optional attributes for BooleanMaskUpdate
Return the shape of s0 op s1 with broadcast.
 
Return the reduction indices for computing gradients of s0 op s1 with broadcast.
 
Broadcast an array for a compatible shape.
 
Bucketizes 'input' based on 'boundaries'.
 
Protobuf type tensorflow.BuildConfiguration
Protobuf type tensorflow.BuildConfiguration
 
Describes the metadata related to a checkpointed tensor.
Describes the metadata related to a checkpointed tensor.
 
Special header that is associated with a bundle.
Special header that is associated with a bundle.
An enum indicating the endianness of the platform that produced this bundle.
 
LINT.IfChange Containers to hold repeated fundamental values.
LINT.IfChange Containers to hold repeated fundamental values.
 
Records the bytes size of each element of input_dataset in a StatsAggregator.
Records the bytes size of each element of input_dataset in a StatsAggregator.
 
 
The CacheDatasetV2 operation
 
Optional attributes for CacheDataset
Defines a subgraph in another `GraphDef` as a set of feed points and nodes to be fetched or executed.
Defines a subgraph in another `GraphDef` as a set of feed points and nodes to be fetched or executed.
 
An n-way switch statement which calls a single branch function.
Optional attributes for Case
Cast x of type SrcT to y of DstT.
 
Optional attributes for Cast
A helper class for casting an Operand
Metric that calculates how often predictions matches one-hot labels.
Computes the crossentropy loss between the labels and predictions.
A Metric that computes the categorical cross-entropy loss between true labels and predicted labels.
Computes the categorical hinge loss between labels and predictions.
A Metric that computes the categorical hinge loss metric between labels and predictions.
Returns element-wise smallest integer not less than x.
 
Checks a tensor for NaN, -Inf and +Inf values.
 
Checks whether a tensor is located in host memory pinned for GPU.
 
Computes the Cholesky decomposition of one or more square matrices.
 
Computes the reverse mode backpropagated gradient of the Cholesky algorithm.
 
The ChooseFastestBranchDataset operation
 
The ChooseFastestDataset operation
The ExperimentalChooseFastestDataset operation
 
 
Clips tensor values to a specified min and max.
 
The CloseSummaryWriter operation
 
Defines a TensorFlow cluster as a set of jobs.
Defines a TensorFlow cluster as a set of jobs.
 
Defines the device filters for jobs in a cluster.
Defines the device filters for jobs in a cluster.
 
An API for building cluster operations as Ops
 
The canonical error codes for TensorFlow APIs.
Code location information: A stack trace with host-name information.
Code location information: A stack trace with host-name information.
 
An op that merges the string-encoded memory config protos from all hosts.
 
CollectionDef should cover most collections.
AnyList is used for collecting Any protos.
AnyList is used for collecting Any protos.
 
CollectionDef should cover most collections.
BytesList is used for collecting strings and serialized protobufs.
BytesList is used for collecting strings and serialized protobufs.
 
FloatList is used for collecting float values.
FloatList is used for collecting float values.
 
Int64List is used for collecting int, int64 and long values.
Int64List is used for collecting int, int64 and long values.
 
 
NodeList is used for collecting nodes in graph.
NodeList is used for collecting nodes in graph.
 
 
Mutually exchanges multiple tensors of identical type and shape.
 
Optional attributes for CollectiveAllToAll
Assign group keys based on group assignment.
 
Receives a tensor value broadcast from another device.
 
Optional attributes for CollectiveBcastRecv
Broadcasts a tensor value to one or more other devices.
 
Optional attributes for CollectiveBcastSend
Mutually accumulates multiple tensors of identical type and shape.
 
Optional attributes for CollectiveGather
Initializes a group for collective operations.
 
Optional attributes for CollectiveInitializeCommunicator
An API for building collective operations as Ops
An Op to permute tensors across replicated TPU instances.
 
Mutually reduces multiple tensors of identical type and shape.
 
Optional attributes for CollectiveReduce
Mutually reduces multiple tensors of identical type and shape and scatters the result.
 
Optional attributes for CollectiveReduceScatter
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.
 
Optional attributes for CombinedNonMaxSuppression
Protobuf type tensorflow.CommitId
Protobuf type tensorflow.CommitId
 
 
Returns the result of a TPU compilation.
 
Compiles a computations for execution on one or more TPU devices.
 
Asserts that compilation succeeded.
 
Converts two real numbers to a complex number.
 
Computes the complex absolute value of a tensor.
 
 
Metadata for CompositeTensorVariant, used when serializing as Variant.
Metadata for CompositeTensorVariant, used when serializing as Variant.
 
Encodes an ExtensionType value into a variant scalar Tensor.
 
Decodes a variant scalar Tensor into an ExtensionType value.
 
Compresses a dataset element.
 
Computes the ids of the positions in sampled_candidates that match true_labels.
 
Optional attributes for ComputeAccidentalHits
Computes the static batch size of a dataset sans partial batches.
 
An op computes the size of the deduplication data from embedding core and returns the updated config.
 
An op computes tuple mask of deduplication data from embedding core.
 
Concatenates tensors along one dimension.
 
Creates a dataset that concatenates input_dataset with another_dataset.
 
Optional attributes for ConcatenateDataset
Concats input tensor across all dimensions.
 
Optional attributes for ConcatND
Computes offsets of concat inputs within its output.
 
A graph that can be invoked as a single function, with an input and output signature.
Protocol buffer representing a CondContext object.
Protocol buffer representing a CondContext object.
 
A conditional accumulator for aggregating gradients.
 
Optional attributes for ConditionalAccumulator
Session configuration parameters.
Session configuration parameters.
Everything inside Experimental is subject to change and is not subject to API stability guarantees in https://www.tensorflow.org/guide/version_compat.
Everything inside Experimental is subject to change and is not subject to API stability guarantees in https://www.tensorflow.org/guide/version_compat.
An enum that describes the state of the MLIR bridge rollout.
 
 
 
An op that sets up the centralized structures for a distributed TPU system.
 
Optional attributes for ConfigureAndInitializeGlobalTPU
Sets up the centralized structures for a distributed TPU system.
 
Optional attributes for ConfigureDistributedTPU
Sets up TPUEmbedding in a distributed TPU system.
 
An op that configures the TPUEmbedding software on a host.
 
An op that configures the TPUEmbedding software on a host.
 
Confusion Matrix Operations
Returns the complex conjugate of a complex number.
 
Shuffle dimensions of x according to a permutation and conjugate the result.
 
An op that sets up communication between TPUEmbedding host software instances after ConfigureTPUEmbeddingHost has been called on each host.
 
Initializer that generates tensors with a constant value.
An operator producing a constant value.
 
This op consumes a lock created by MutexLock.
 
Container for any kind of control flow context.
Container for any kind of control flow context.
 
 
 
Does nothing.
 
Computes a N-D convolution given (N+1+batch_dims)-D input and (N+2)-D filter tensors.
 
Optional attributes for Conv
Computes a 2-D convolution given 4-D input and filter tensors.
 
Optional attributes for Conv2d
Computes the gradients of convolution with respect to the filter.
 
Optional attributes for Conv2dBackpropFilter
Computes the gradients of convolution with respect to the filter.
 
Optional attributes for Conv2dBackpropFilterV2
Computes the gradients of convolution with respect to the input.
 
Optional attributes for Conv2dBackpropInput
Computes the gradients of convolution with respect to the input.
 
Optional attributes for Conv2dBackpropInputV2
Computes a 3-D convolution given 5-D input and filter tensors.
 
Optional attributes for Conv3d
Computes the gradients of 3-D convolution with respect to the filter.
 
Optional attributes for Conv3dBackpropFilter
Computes the gradients of 3-D convolution with respect to the input.
 
Optional attributes for Conv3dBackpropInput
The ConvertToCooTensor operation
 
The ConvertToListOfSparseCoreCooTensors operation
 
The ConvertToSparseCoreCsrWrappedCooTensor operation
 
 
Represents a job type and the number of tasks under this job.
Represents a job type and the number of tasks under this job.
 
Coordination service configuration parameters.
Coordination service configuration parameters.
 
Copy a tensor from CPU-to-CPU or GPU-to-GPU.
 
Optional attributes for Copy
Copy a tensor to host.
 
Optional attributes for CopyHost
The CopyToMesh operation
 
The CopyToMeshGrad operation
 
 
If included as a payload, this message contains the error source information where the error was raised.
If included as a payload, this message contains the error source information where the error was raised.
Protobuf enum tensorflow.core.platform.ErrorSourceProto.ErrorSource
 
Computes cos of x element-wise.
 
Computes hyperbolic cosine of x element-wise.
 
Computes the cosine similarity between labels and predictions.
A metric that computes the cosine similarity metric between labels and predictions.
Protobuf type tensorflow.CostGraphDef
Total cost of this graph, typically used for balancing decisions.
Total cost of this graph, typically used for balancing decisions.
 
Protobuf type tensorflow.CostGraphDef
Protobuf type tensorflow.CostGraphDef.Node
Protobuf type tensorflow.CostGraphDef.Node
Inputs of this node.
Inputs of this node.
 
Outputs of this node.
Outputs of this node.
 
 
 
 
Increments 'ref' until it reaches 'limit'.
 
 
Protobuf type tensorflow.core.CppShapeInferenceInputsNeeded
Protobuf type tensorflow.core.CppShapeInferenceInputsNeeded
 
Protobuf type tensorflow.core.CppShapeInferenceResult
Protobuf type tensorflow.core.CppShapeInferenceResult
Protobuf type tensorflow.core.CppShapeInferenceResult.HandleData
Protobuf type tensorflow.core.CppShapeInferenceResult.HandleData
 
Protobuf type tensorflow.core.CppShapeInferenceResult.HandleShapeAndType
Protobuf type tensorflow.core.CppShapeInferenceResult.HandleShapeAndType
 
 
Protobuf type tensorflow.CPUInfo
Protobuf type tensorflow.CPUInfo
 
The CreateSummaryDbWriter operation
 
The CreateSummaryFileWriter operation
 
Extracts crops from the input image tensor and resizes them.
 
Optional attributes for CropAndResize
Computes the gradient of the crop_and_resize op wrt the input boxes tensor.
 
Optional attributes for CropAndResizeGradBoxes
Computes the gradient of the crop_and_resize op wrt the input image tensor.
 
Optional attributes for CropAndResizeGradImage
Compute the pairwise cross product.
 
An Op to sum inputs across replicated TPU instances.
 
Reads out the CSR components at batch index.
 
Convert a (possibly batched) CSRSparseMatrix to dense.
 
Converts a (possibly batched) CSRSparesMatrix to a SparseTensor.
 
The CSVDatasetV2 operation
The ExperimentalCSVDataset operation
 
 
Performs beam search decoding on the logits given in input.
 
Optional attributes for CtcBeamSearchDecoder
Performs greedy decoding on the logits given in inputs.
 
Optional attributes for CtcGreedyDecoder
Calculates the CTC Loss (log probability) for each batch entry.
 
Optional attributes for CtcLoss
Calculates the CTC Loss (log probability) for each batch entry.
 
Optional attributes for CTCLossV2
A RNN backed by cuDNN.
 
Optional attributes for CudnnRNN
Backprop step of CudnnRNNV3.
 
Optional attributes for CudnnRNNBackprop
Converts CudnnRNN params from canonical form to usable form.
 
Optional attributes for CudnnRNNCanonicalToParams
Computes size of weights that can be used by a Cudnn RNN model.
 
Optional attributes for CudnnRnnParamsSize
Retrieves CudnnRNN params in canonical form.
 
Optional attributes for CudnnRNNParamsToCanonical
Compute the cumulative product of the tensor x along axis.
 
Optional attributes for Cumprod
Compute the cumulative sum of the tensor x along axis.
 
Optional attributes for Cumsum
Compute the cumulative product of the tensor x along axis.
 
Optional attributes for CumulativeLogsumexp
A custom gradient for ops of type T.
Protobuf enum tensorflow.DataClass
An API for building data.experimental operations as Ops
Returns the dimension index in the destination data format given the one in the source data format.
 
Optional attributes for DataFormatDimMap
Permute input tensor from src_format to dst_format.
 
Optional attributes for DataFormatVecPermute
An API for building data operations as Ops
 
Protobuf type tensorflow.data.CrossTrainerCacheOptions
Protobuf type tensorflow.data.CrossTrainerCacheOptions
 
Data service config available to the client through GetDataServiceConfig RPC.
Data service config available to the client through GetDataServiceConfig RPC.
 
Metadata related to tf.data service datasets.
Metadata related to tf.data service datasets.
Protobuf enum tensorflow.data.DataServiceMetadata.Compression
 
 
tf.data service deployment mode.
Next tag: 2
Next tag: 2
Specifies how data is sharded among tf.data service workers.
 
Creates a dataset that reads data from the tf.data service.
 
Optional attributes for DataServiceDataset
Represents a potentially large list of independent elements (samples), and allows iteration and transformations to be performed across these elements.
 
Metadata describing a compressed component of a dataset element.
Metadata describing a compressed component of a dataset element.
 
Protobuf type tensorflow.data.CompressedElement
Protobuf type tensorflow.data.CompressedElement
 
An uncompressed dataset element.
An uncompressed dataset element.
 
Returns the cardinality of input_dataset.
Returns the cardinality of input_dataset.
 
 
Optional attributes for DatasetCardinality
Returns the fingerprint of input_dataset.
 
Creates a dataset from the given graph_def.
 
Represents the state of an iteration through a tf.data Datset.
 
next: 2
next: 2
 
An optional represents the result of a dataset getNext operation that may fail, when the end of the dataset has been reached.
 
Represents the type of auto-sharding we enable.
next: 6
next: 6
 
 
 
 
 
 
next: 2
next: 2
Protobuf enum tensorflow.data.CardinalityOptions.ComputeLevel
 
next: 3
next: 3
 
 
Represents how to handle external state during serialization.
next: 22
next: 22
 
 
 
 
 
 
 
 
 
 
 
 
 
Message stored with Dataset objects to control how datasets are processed and optimized. next: 13
Message stored with Dataset objects to control how datasets are processed and optimized. next: 13
 
 
 
 
 
 
 
next: 2
next: 2
 
 
next: 3
next: 3
 
 
 
Returns a serialized GraphDef representing input_dataset.
 
Optional attributes for DatasetToGraph
Outputs the single element from the given dataset.
 
Optional attributes for DatasetToSingleElement
Writes the given dataset to the given file using the TFRecord format.
Writes the given dataset to the given file using the TFRecord format.
 
 
(== suppress_warning documentation-presence ==) LINT.IfChange
The Dawsn operation
 
 
An Event related to the debugging of a TensorFlow program.
An Event related to the debugging of a TensorFlow program.
 
 
 
A device on which ops and/or tensors are instrumented by the debugger.
A device on which ops and/or tensors are instrumented by the debugger.
 
A debugger-instrumented graph.
A debugger-instrumented graph.
 
Protobuf type tensorflow.DebuggedSourceFile
Protobuf type tensorflow.DebuggedSourceFile
 
Protobuf type tensorflow.DebuggedSourceFiles
Protobuf type tensorflow.DebuggedSourceFiles
 
An API for building debugging operations as Ops
Identity op for gradient debugging.
 
Identity op for gradient debugging.
 
Provides an identity mapping of the non-Ref type input tensor for debugging.
 
Optional attributes for DebugIdentity
Metadata about the debugger and the debugged TensorFlow program.
Metadata about the debugger and the debugged TensorFlow program.
 
Debug NaN Value Counter Op.
 
Optional attributes for DebugNanCount
Debug Numeric Summary V2 Op.
 
Optional attributes for DebugNumericsSummary
Options for initializing DebuggerState in TensorFlow Debugger (tfdbg).
Options for initializing DebuggerState in TensorFlow Debugger (tfdbg).
 
 
Option for watching a node in TensorFlow Debugger (tfdbg).
Option for watching a node in TensorFlow Debugger (tfdbg).
 
Decode and Crop a JPEG-encoded image to a uint8 tensor.
 
Optional attributes for DecodeAndCropJpeg
Decode web-safe base64-encoded strings.
 
Decode the first frame of a BMP-encoded image to a uint8 tensor.
 
Optional attributes for DecodeBmp
Decompress strings.
 
Optional attributes for DecodeCompressed
Convert CSV records to tensors.
 
Optional attributes for DecodeCsv
Decode the frame(s) of a GIF-encoded image to a uint8 tensor.
 
Function for decode_bmp, decode_gif, decode_jpeg, and decode_png.
 
Optional attributes for DecodeImage
Decode a JPEG-encoded image to a uint8 tensor.
 
Optional attributes for DecodeJpeg
Convert JSON-encoded Example records to binary protocol buffer strings.
 
Reinterpret the bytes of a string as a vector of numbers.
 
Optional attributes for DecodePaddedRaw
Decode a PNG-encoded image to a uint8 or uint16 tensor.
 
Optional attributes for DecodePng
The op extracts fields from a serialized protocol buffers message into tensors.
 
Optional attributes for DecodeProto
Reinterpret the bytes of a string as a vector of numbers.
 
Optional attributes for DecodeRaw
Decode a 16-bit PCM WAV file to a float tensor.
 
Optional attributes for DecodeWav
Makes a copy of x.
 
A container for an iterator resource.
 
The DeleteMemoryCache operation
 
A container for an iterator resource.
 
The DeleteRandomSeedGenerator operation
 
The DeleteSeedGenerator operation
 
Delete the tensor specified by its handle in the session.
 
Counts the number of occurrences of each value in an integer array.
 
Optional attributes for DenseBincount
Performs sparse-output bin counting for a tf.tensor input.
 
Optional attributes for DenseCountSparseOutput
Converts a dense tensor to a (possibly batched) CSRSparseMatrix.
 
Applies set operation along last dimension of 2 Tensor inputs.
 
Optional attributes for DenseToDenseSetOperation
Creates a dataset that batches input elements into a SparseTensor.
Creates a dataset that batches input elements into a SparseTensor.
 
 
Applies set operation along last dimension of Tensor and SparseTensor.
 
Optional attributes for DenseToSparseSetOperation
DepthToSpace for tensors of type T.
 
Optional attributes for DepthToSpace
Computes a 2-D depthwise convolution given 4-D input and filter tensors.
 
Optional attributes for DepthwiseConv2dNative
Computes the gradients of depthwise convolution with respect to the filter.
 
Optional attributes for DepthwiseConv2dNativeBackpropFilter
Computes the gradients of depthwise convolution with respect to the input.
 
Optional attributes for DepthwiseConv2dNativeBackpropInput
Dequantize the 'input' tensor into a float or bfloat16 Tensor.
 
Optional attributes for Dequantize
Converts the given variant tensor to an iterator and stores it in the given resource.
 
Deserialize and concatenate SparseTensors from a serialized minibatch.
 
Deserialize SparseTensor objects.
 
Deletes the resource specified by the handle.
 
Optional attributes for DestroyResourceOp
Destroys the temporary variable and returns its final value.
 
Computes the determinant of one or more square matrices.
 
Protobuf type tensorflow.DeviceAttributes
Protobuf type tensorflow.DeviceAttributes
 
 
 
Return the index of device the op runs.
 
Protobuf type tensorflow.DeviceLocality
Protobuf type tensorflow.DeviceLocality
 
 
Protobuf type tensorflow.DeviceProperties
Protobuf type tensorflow.DeviceProperties
 
Protobuf type tensorflow.NamedDevice
Protobuf type tensorflow.NamedDevice
 
Represents a (possibly partial) specification for a TensorFlow device.
A Builder class for building DeviceSpec class.
 
Protobuf type tensorflow.DeviceStepStats
Protobuf type tensorflow.DeviceStepStats
 
Computes Psi, the derivative of Lgamma (the log of the absolute value of Gamma(x)), element-wise.
 
Computes the grayscale dilation of 4-D input and 3-D filter tensors.
 
Computes the gradient of morphological 2-D dilation with respect to the filter.
 
Computes the gradient of morphological 2-D dilation with respect to the input.
 
A substitute for InterleaveDataset on a fixed list of N datasets.
A substitute for InterleaveDataset on a fixed list of N datasets.
 
 
Optional attributes for DirectedInterleaveDataset
Turns off the copy-on-read mode.
 
 
Used to serialize and transmit tensorflow::Status payloads through grpc::Status `error_details` since grpc::Status lacks payload API.
Used to serialize and transmit tensorflow::Status payloads through grpc::Status `error_details` since grpc::Status lacks payload API.
 
If included as a payload, this message flags the Status to have lost payloads during the GRPC transmission.
If included as a payload, this message flags the Status to have lost payloads during the GRPC transmission.
 
If included as a payload, this message flags the Status to be a possible outcome of a worker restart.
If included as a payload, this message flags the Status to be a possible outcome of a worker restart.
 
The DistributedSave operation
 
Optional attributes for DistributedSave
An API for building distribute operations as Ops
Returns x / y element-wise.
 
Returns 0 if the denominator is zero.
 
Draw bounding boxes on a batch of images.
 
The DTensorRestoreV2 operation
 
An API for building dtypes operations as Ops
The DummyIterationCounter operation
 
The DummyMemoryCache operation
 
The DummySeedGenerator operation
 
Eases the porting of code that uses tf.nn.embedding_lookup_sparse(). embedding_indices[i] and aggregation_weights[i] correspond to the ith feature.
 
The DynamicEnqueueTPUEmbeddingRaggedTensorBatch operation
 
Partitions data into num_partitions tensors using indices from partitions.
 
Interleave the values from the data tensors into a single tensor.
 
An environment for executing TensorFlow operations eagerly.
Controls how to act when we try to run an operation on a given device but some input tensors are not on that device.
 
Computes the (possibly normalized) Levenshtein Edit Distance.
 
Optional attributes for EditDistance
Computes the eigen decomposition of one or more square matrices.
 
Optional attributes for Eig
Tensor contraction according to Einstein summation convention.
 
Computes the exponential linear function.
Exponential linear unit.
 
Computes gradients for the exponential linear (Elu) operation.
 
An op enabling differentiation of TPU Embeddings.
 
Creates a tensor with the given shape.
 
Optional attributes for Empty
Creates and returns an empty tensor list.
 
Creates and returns an empty tensor map. handle: an empty tensor map
 
Encode strings into web-safe base64 format.
 
Optional attributes for EncodeBase64
JPEG-encode an image.
 
Optional attributes for EncodeJpeg
JPEG encode input image with provided compression quality.
 
PNG-encode an image.
 
Optional attributes for EncodePng
The op serializes protobuf messages provided in the input tensors.
 
Optional attributes for EncodeProto
Encode audio data using the WAV file format.
 
Annotation used to mark a method of a class annotated with @Operator that should generate an endpoint into Ops or one of its groups.
Eases the porting of code that uses tf.nn.embedding_lookup_sparse(). embedding_indices[i] and aggregation_weights[i] correspond to the ith feature.
 
An op that enqueues a list of input batch tensors to TPUEmbedding.
 
Optional attributes for EnqueueTPUEmbeddingBatch
An op that enqueues a list of input batch tensors to TPUEmbedding.
 
Optional attributes for EnqueueTPUEmbeddingIntegerBatch
Eases the porting of code that uses tf.nn.embedding_lookup(). sample_splits[i], embedding_indices[i] and aggregation_weights[i] correspond to the ith feature. table_ids[i] indicates which embedding table to look up ith feature.
 
Optional attributes for EnqueueTPUEmbeddingRaggedTensorBatch
An op that enqueues TPUEmbedding input indices from a SparseTensor.
 
Optional attributes for EnqueueTPUEmbeddingSparseBatch
Eases the porting of code that uses tf.nn.embedding_lookup_sparse(). sample_indices[i], embedding_indices[i] and aggregation_weights[i] correspond to the ith feature. table_ids[i] indicates which embedding table to look up ith feature.
 
Optional attributes for EnqueueTPUEmbeddingSparseTensorBatch
Ensures that the tensor's shape matches the expected shape.
 
Creates or finds a child frame, and makes data available to the child frame.
 
Optional attributes for Enter
Protobuf type tensorflow.EntryValue
Protobuf type tensorflow.EntryValue
 
 
Returns the truth value of (x == y) element-wise.
 
Optional attributes for Equal
Computes the Gauss error function of x element-wise.
 
Computes the complementary error function of x element-wise.
 
The Erfinv operation
 
 
 
 
Computes the euclidean norm of elements across dimensions of a tensor.
 
Optional attributes for EuclideanNorm
Protocol buffer representing an event that happened during the execution of a Brain model.
Protocol buffer representing an event that happened during the execution of a Brain model.
 
 
 
Protobuf type tensorflow.Example
Protobuf type tensorflow.Example
 
Protobuf type tensorflow.ExampleParserConfiguration
Protobuf type tensorflow.ExampleParserConfiguration
 
 
 
Op that loads and executes a TPU program on a TPU device.
 
Op that executes a program with optional in-place variable updates.
 
An op that executes the TPUEmbedding partitioner on the central configuration device and computes the HBM size (in bytes) required for TPUEmbedding operation.
 
Data relating to the eager execution of an op or a Graph.
Data relating to the eager execution of an op or a Graph.
Defines an environment for creating and executing TensorFlow Operations.
 
 
Exits the current frame to its parent frame.
 
Computes exponential of x element-wise.
 
Inserts a dimension of 1 into a tensor's shape.
 
The Expint operation
 
Computes exp(x) - 1 element-wise. i.e.
 
Exponential activation function.
Extracts a glimpse from the input tensor.
 
Optional attributes for ExtractGlimpse
Extract patches from images and put them in the "depth" output dimension.
 
Extract the shape information of a JPEG-encoded image.
 
Extract patches from input and put them in the "depth" output dimension. 3D extension of extract_image_patches.
 
Output a fact about factorials.
 
This op is used as a placeholder in If branch functions.
 
Fake-quantize the 'inputs' tensor, type float to 'outputs' tensor of same shape and type.
 
Optional attributes for FakeQuantWithMinMaxArgs
Compute gradients for a FakeQuantWithMinMaxArgs operation.
 
Optional attributes for FakeQuantWithMinMaxArgsGradient
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.
 
Optional attributes for FakeQuantWithMinMaxVars
Compute gradients for a FakeQuantWithMinMaxVars operation.
 
Optional attributes for FakeQuantWithMinMaxVarsGradient
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.
 
Optional attributes for FakeQuantWithMinMaxVarsPerChannel
Compute gradients for a FakeQuantWithMinMaxVarsPerChannel operation.
 
Deprecated.
 
Metric that calculates the number of false negatives.
Metric that calculates the number of false positives.
Containers for non-sequential data.
Containers for non-sequential data.
 
Protobuf type tensorflow.FeatureConfiguration
Protobuf type tensorflow.FeatureConfiguration
 
 
Containers for sequential data.
Containers for sequential data.
 
Protobuf type tensorflow.FeatureLists
Protobuf type tensorflow.FeatureLists
 
 
 
Protobuf type tensorflow.Features
Protobuf type tensorflow.Features
 
Fast Fourier transform.
 
2D fast Fourier transform.
 
3D fast Fourier transform.
 
ND fast Fourier transform.
 
A queue that produces elements in first-in first-out order.
 
Optional attributes for FifoQueue
Set configuration of the file system.
 
Creates a tensor filled with a scalar value.
 
Creates a dataset containing elements of first component of input_dataset having true in the last component.
 
Creates a dataset containing elements of input_dataset matching predicate.
 
Optional attributes for FilterDataset
Creates a dataset by applying tf.data.Options to input_dataset.
 
Optional attributes for FinalizeDataset
An op that finalizes the TPUEmbedding configuration.
 
Generates fingerprint values.
 
Protocol buffer representing a SavedModel Fingerprint.
Protocol buffer representing a SavedModel Fingerprint.
 
 
Protobuf type tensorflow.FixedLenFeatureProto
Protobuf type tensorflow.FixedLenFeatureProto
 
The FixedLengthRecordDatasetV2 operation
 
Optional attributes for FixedLengthRecordDataset
A Reader that outputs fixed-length records from a file.
 
Optional attributes for FixedLengthRecordReader
Generates labels for candidate sampling with a learned unigram distribution.
 
Optional attributes for FixedUnigramCandidateSampler
Creates a dataset that applies f to the outputs of input_dataset.
 
Optional attributes for FlatMapDataset
Protobuf type tensorflow.FloatList
Protobuf type tensorflow.FloatList
 
Returns element-wise largest integer not greater than x.
 
Returns x // y element-wise.
 
Returns element-wise remainder of division.
 
The FlushSummaryWriter operation
 
Applies a for loop.
 
Performs fractional average pooling on the input.
 
Optional attributes for FractionalAvgPool
Computes gradient of the FractionalAvgPool function.
 
Optional attributes for FractionalAvgPoolGrad
Performs fractional max pooling on the input.
 
Optional attributes for FractionalMaxPool
Computes gradient of the FractionalMaxPool function.
 
Optional attributes for FractionalMaxPoolGrad
An API for building framework operations as Ops
The FresnelCos operation
 
The FresnelSin operation
 
Optimizer that implements the FTRL algorithm.
Highly experimental and very likely to change.
 
Highly experimental and very likely to change.
 
LINT.IfChange Experimental.
 
Ops for calling ConcreteFunction.
A function can be instantiated when the runtime can bind every attr with a value.
Attributes for function arguments.
Attributes for function arguments.
 
A function can be instantiated when the runtime can bind every attr with a value.
A library is a set of named functions.
A library is a set of named functions.
 
 
 
Batch normalization.
 
Optional attributes for FusedBatchNorm
Gradient for batch normalization.
 
Optional attributes for FusedBatchNormGrad
Performs a padding as a preprocess during a convolution.
 
Performs a resize and padding as a preprocess during a convolution.
 
Optional attributes for FusedResizeAndPadConv2d
Gather slices from params axis axis according to indices.
 
Optional attributes for Gather
Gather slices from params into a Tensor with shape specified by indices.
 
Optional attributes for GatherNd
Applies the Gaussian error linear unit (GELU) activation function.
The Gaussian Error Linear Unit (GELU) activation function.
This op produces Region of Interests from given bounding boxes(bbox_deltas) encoded wrt anchors according to eq.2 in arXiv:1506.01497
 
Optional attributes for GenerateBoundingBoxProposals
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.
 
Optional attributes for GenerateVocabRemapping
Creates a dataset that invokes a function to generate elements.
 
Optional attributes for GeneratorDataset
Gets the element at the specified index in a dataset.
 
The GetMinibatchesInCsrWithPhysicalReplica operation
 
The GetMinibatchSplitsWithPhysicalReplica operation
 
Returns the tf.data.Options attached to input_dataset.
 
Store the input tensor in the state of the current session.
 
Get the value of the tensor specified by its handle.
 
The GetStatsFromListOfSparseCoreCooTensors operation
 
An op returns the TPU task ID from TPU topology.
 
The GlobalIterId operation
 
The GlobalShuffleDataset operation
 
Optional attributes for GlobalShuffleDataset
The Glorot initializer, also called Xavier initializer.
Protobuf type tensorflow.GPUInfo
Protobuf type tensorflow.GPUInfo
 
Protobuf type tensorflow.GPUOptions
Protobuf type tensorflow.GPUOptions
Protobuf type tensorflow.GPUOptions.Experimental
Protobuf type tensorflow.GPUOptions.Experimental
Whether to merge data transfer streams into the compute stream in the same stream group.
Whether to merge data transfer streams into the compute stream in the same stream group.
 
Configuration for breaking down a visible GPU into multiple "virtual" devices.
Configuration for breaking down a visible GPU into multiple "virtual" devices.
 
 
 
GradientDef defines the gradient function of a function defined in a function library.
GradientDef defines the gradient function of a function defined in a function library.
 
Basic Stochastic gradient descent optimizer.
Adds operations to compute the partial derivatives of sum of ys w.r.t xs, i.e., d(y_1 + y_2 + ...)/dx_1, d(y_1 + y_2 + ...)/dx_2...
Optional attributes for Gradients
A data flow graph representing a TensorFlow computation.
Used to instantiate an abstract class which overrides the buildSubgraph method to build a conditional or body subgraph for a while loop.
Protobuf type tensorflow.GraphDebugInfo
Protobuf type tensorflow.GraphDebugInfo
This represents a file/line location in the source code.
This represents a file/line location in the source code.
 
This represents a stack trace which is a ordered list of `FileLineCol`.
This represents a stack trace which is a ordered list of `FileLineCol`.
 
 
 
Represents the graph of operations
Represents the graph of operations
 
Data relating to an execution of a Graph (e.g., an eager execution of a FuncGraph).
Data relating to an execution of a Graph (e.g., an eager execution of a FuncGraph).
 
The creation of an op in a TensorFlow Graph (e.g., FuncGraph in TF2).
The creation of an op in a TensorFlow Graph (e.g., FuncGraph in TF2).
 
Implementation for an Operation added as a node to a Graph.
Protobuf type tensorflow.GraphOptions
Protobuf type tensorflow.GraphOptions
 
 
Protobuf type tensorflow.GraphTransferConstNodeInfo
Protobuf type tensorflow.GraphTransferConstNodeInfo
 
Protobuf type tensorflow.GraphTransferGraphInputNodeInfo
Protobuf type tensorflow.GraphTransferGraphInputNodeInfo
 
Protobuf type tensorflow.GraphTransferGraphOutputNodeInfo
Protobuf type tensorflow.GraphTransferGraphOutputNodeInfo
 
Protocol buffer representing a handle to a tensorflow resource.
Protocol buffer representing a handle to a tensorflow resource.
Protobuf enum tensorflow.GraphTransferInfo.Destination
 
 
Protobuf type tensorflow.GraphTransferNodeInfo
Protobuf type tensorflow.GraphTransferNodeInfo
 
Protobuf type tensorflow.GraphTransferNodeInput
Protobuf type tensorflow.GraphTransferNodeInput
Protobuf type tensorflow.GraphTransferNodeInputInfo
Protobuf type tensorflow.GraphTransferNodeInputInfo
 
 
Protobuf type tensorflow.GraphTransferNodeOutputInfo
Protobuf type tensorflow.GraphTransferNodeOutputInfo
 
Returns the truth value of (x > y) element-wise.
 
Returns the truth value of (x >= y) element-wise.
 
Creates a dataset that computes a group-by on input_dataset.
Creates a dataset that computes a group-by on input_dataset.
 
 
Creates a dataset that computes a windowed group-by on input_dataset. // TODO(mrry): Support non-int64 keys.
Creates a dataset that computes a windowed group-by on input_dataset. // TODO(mrry): Support non-int64 keys.
 
 
Optional attributes for GroupByWindowDataset
Computes the GRU cell forward propagation for 1 time step.
 
Computes the GRU cell back-propagation for 1 time step.
 
Gives a guarantee to the TF runtime that the input tensor is a constant.
 
Hard sigmoid activation.
Creates a non-initialized hash table.
 
Optional attributes for HashTable
He initializer.
Container class for core methods which add or perform several operations and return one of them.
Computes the hinge loss between labels and predictions.
A metric that computes the hinge loss metric between labels and predictions.
 
Return histogram of values.
 
Serialization format for histogram module in tsl/lib/histogram/histogram.h
Serialization format for histogram module in tsl/lib/histogram/histogram.h
 
Outputs a Summary protocol buffer with a histogram.
 
Returns a constant tensor on the host.
 
Convert one or more images from HSV to RGB.
 
Computes the Huber loss between labels and predictions.
Initializer that generates the identity matrix.
Return a tensor with the same shape and contents as the input tensor or value.
 
Returns a list of tensors with the same shapes and contents as the input tensors.
 
A Reader that outputs the queued work as both the key and value.
 
Optional attributes for IdentityReader
output = cond ?
Optional attributes for If
Inverse fast Fourier transform.
 
Inverse 2D fast Fourier transform.
 
Inverse 3D fast Fourier transform.
 
ND inverse fast Fourier transform.
 
Compute the lower regularized incomplete Gamma function P(a, x).
 
Compute the upper regularized incomplete Gamma function Q(a, x).
 
Computes the gradient of igamma(a, x) wrt a.
 
Creates a dataset that contains the elements of input_dataset ignoring errors.
Creates a dataset that contains the elements of input_dataset ignoring errors.
 
 
Optional attributes for IgnoreErrorsDataset
Optional attributes for IgnoreErrorsDataset
Returns the imaginary part of a complex number.
 
An API for building image operations as Ops
Applies the given transform to each of the images.
 
Optional attributes for ImageProjectiveTransformV2
Applies the given transform to each of the images.
 
Optional attributes for ImageProjectiveTransformV3
Outputs a Summary protocol buffer with images.
 
Optional attributes for ImageSummary
Returns immutable tensor from memory region.
 
The ImportEvent operation
 
The IndexFlatMapDataset operation
 
Optional attributes for IndexFlatMapDataset
A placeholder op for a value that will be fed into the computation.
 
Fetches multiple values from infeed as an XLA tuple.
 
An op which feeds a single Tensor value into the computation.
 
Optional attributes for InfeedEnqueue
An op which enqueues prelinearized buffer into TPU infeed.
 
Optional attributes for InfeedEnqueuePrelinearizedBuffer
Feeds multiple Tensor values into the computation as an XLA tuple.
 
Optional attributes for InfeedEnqueueTuple
An interface for Initializers
Table initializer that takes two tensors for keys and values respectively.
 
The InitializeTableFromDataset operation
 
Initializes a table from a text file.
 
Optional attributes for InitializeTableFromTextFile
Adds v into specified rows of x.
 
Subtracts `v` into specified rows of `x`.
 
Updates specified rows 'i' with values 'v'.
 
Protobuf type tensorflow.Int64List
Protobuf type tensorflow.Int64List
 
Protobuf type tensorflow.InterconnectLink
Protobuf type tensorflow.InterconnectLink
 
Creates a dataset that applies f to the outputs of input_dataset.
 
Optional attributes for InterleaveDataset
Says whether the targets are in the top K predictions.
 
Computes the inverse of one or more square invertible matrices or their adjoints (conjugate transposes).
 
Optional attributes for Inv
Invert (flip) each bit of supported types; for example, type uint8 value 01010101 becomes 10101010.
 
Computes the inverse permutation of a tensor.
 
Computes the gradient for the inverse of x wrt its input.
 
An API for building io operations as Ops
Inverse real-valued fast Fourier transform.
 
Inverse 2D real-valued fast Fourier transform.
 
Inverse 3D real-valued fast Fourier transform.
 
ND inverse real fast Fourier transform.
 
Returns which elements of x are finite.
 
Returns which elements of x are Inf.
 
Returns which elements of x are NaN.
 
Solves a batch of isotonic regression problems.
 
Whether TPU Embedding is initialized in a distributed TPU system.
 
Optional attributes for IsTPUEmbeddingInitialized
Checks whether a tensor has been initialized.
 
The IteratorV2 operation
 
The IteratorFromStringHandleV2 operation
 
Optional attributes for IteratorFromStringHandle
Returns the name of the device on which resource has been placed.
Returns the name of the device on which resource has been placed.
 
 
Returns the serialized model proto of an iterator resource.
 
Gets the next output from the given iterator .
 
Gets the next output from the given iterator as an Optional variant.
 
Gets the next output from the given iterator.
 
Converts the given resource_handle representing an iterator to a string.
 
Defines a single job in a TensorFlow cluster.
Defines a single job in a TensorFlow cluster.
 
Defines the device filters for tasks in a job.
Defines the device filters for tasks in a job.
 
Joins the strings in the given list of string tensors into one tensor; with the given separator (default is an empty separator).
 
Optional attributes for Join
Protobuf type tensorflow.KernelDef
Protobuf type tensorflow.KernelDef.AttrConstraint
Protobuf type tensorflow.KernelDef.AttrConstraint
 
Protobuf type tensorflow.KernelDef
 
 
A collection of KernelDefs
A collection of KernelDefs
 
Computes Kullback-Leibler divergence loss between labels and predictions.
A metric that computes the Kullback-Leibler divergence loss metric between labels and predictions.
Returns the index of a data point that should be added to the seed set.
 
Selects num_to_sample rows of input using the KMeans++ criterion.
 
Computes the Kth order statistic of a data set.
 
A regularizer that applies an L1 or Lasso(least absolute shrinkage and selection operator) Regression, regularization penalty.
A regularizer that applies both L1 and L2 regularization penalties.
A regularizer that applies a L2 (Ridge Regression) regularization penalty.
L2 Loss.
 
L2 Normalization Operations
Records the latency of producing input_dataset elements in a StatsAggregator.
Records the latency of producing input_dataset elements in a StatsAggregator.
 
 
Computes rectified linear: max(features, features * alpha).
 
Optional attributes for LeakyRelu
Computes rectified linear gradients for a LeakyRelu operation.
 
Optional attributes for LeakyReluGrad
Generates labels for candidate sampling with a learned unigram distribution.
 
Optional attributes for LearnedUnigramCandidateSampler
LeCun normal initializer.
Elementwise computes the bitwise left-shift of x and y.
 
Creates a dataset that applies f to the outputs of input_dataset.
 
Optional attributes for LegacyParallelInterleaveDataset
Returns the truth value of (x < y) element-wise.
 
Returns the truth value of (x <= y) element-wise.
 
Computes the log of the absolute value of Gamma(x) element-wise.
 
 
An API for building linalg operations as Ops
An API for building linalg.sparse operations as Ops
Linear activation function (pass-through).
Generates values in an interval.
 
Creates a dataset that emits each of tensors once.
 
Optional attributes for ListDataset
 
 
The ListSnapshotChunksDataset operation
 
The ExperimentalLMDBDataset operation
Creates a dataset that emits the key-value pairs in one or more LMDB files.
 
 
A Reader that outputs the records from a LMDB file.
 
Optional attributes for LmdbReader
An op that loads optimization parameters into embedding memory.
 
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.
 
Optional attributes for LoadAndRemapMatrix
The LoadDataset operation
 
Optional attributes for LoadDataset
Load Adadelta embedding parameters.
 
Optional attributes for LoadTPUEmbeddingAdadeltaParameters
Load Adagrad Momentum embedding parameters.
 
Load Adagrad embedding parameters.
 
Optional attributes for LoadTPUEmbeddingAdagradParameters
Load ADAM embedding parameters.
 
Optional attributes for LoadTPUEmbeddingADAMParameters
Load centered RMSProp embedding parameters.
 
Load frequency estimator embedding parameters.
 
Load FTRL embedding parameters.
 
Optional attributes for LoadTPUEmbeddingFTRLParameters
Load MDL Adagrad Light embedding parameters.
 
Load Momentum embedding parameters.
 
Optional attributes for LoadTPUEmbeddingMomentumParameters
Load proximal Adagrad embedding parameters.
 
The LoadTPUEmbeddingProximalYogiParameters operation
 
Load RMSProp embedding parameters.
 
Optional attributes for LoadTPUEmbeddingRMSPropParameters
Load SGD embedding parameters.
 
Protobuf type tensorflow.LocalLinks
Protobuf type tensorflow.LocalLinks
 
Local Response Normalization.
 
Optional attributes for LocalResponseNormalization
Gradients for Local Response Normalization.
 
Optional attributes for LocalResponseNormalizationGrad
Computes natural logarithm of x element-wise.
 
Computes natural logarithm of (1 + x) element-wise.
 
Computes Computes the logarithm of the hyperbolic cosine of the prediction error.
A metric that computes the logarithm of the hyperbolic cosine of the prediction error metric between labels and predictions.
Returns the truth value of x AND y element-wise.
 
Returns the truth value of NOT x element-wise.
 
Returns the truth value of x OR y element-wise.
 
Computes the sign and the log of the absolute value of the determinant of one or more square matrices.
 
 
Deprecated.
Protocol buffer used for logging messages to the events file.
Deprecated.
Deprecated.
Computes log softmax activations.
 
Generates labels for candidate sampling with a log-uniform distribution.
 
Optional attributes for LogUniformCandidateSampler
Outputs all keys and values in the table.
 
Looks up keys in a table, outputs the corresponding values.
 
Replaces the contents of the table with the specified keys and values.
 
Updates the table to associates keys with values.
 
Removes keys and its associated values from a table.
 
Computes the number of elements in the given table.
 
Forwards the input to the output.
 
Interface for loss calc ulation
Built-in loss functions.
Converts all uppercase characters into their respective lowercase replacements.
 
Optional attributes for Lower
Applies lower_bound(sorted_search_values, values) along each row.
 
Computes the LSTM cell forward propagation for 1 time step.
 
Optional attributes for LSTMBlockCell
Computes the LSTM cell backward propagation for 1 timestep.
 
Computes the LU decomposition of one or more square matrices.
 
Protobuf type tensorflow.MachineConfiguration
Protobuf type tensorflow.MachineConfiguration
 
Makes a new iterator from the given dataset and stores it in iterator.
 
Make all elements in the non-Batch dimension unique, but "close" to their initial value.
 
Creates a dataset that fuses mapping with batching.
Creates a dataset that fuses mapping with batching.
 
 
Optional attributes for MapAndBatchDataset
Optional attributes for MapAndBatchDataset
Op removes all elements in the underlying container.
 
Optional attributes for MapClear
Creates a dataset that applies f to the outputs of input_dataset.
Creates a dataset that applies f to the outputs of input_dataset.
 
 
Optional attributes for MapDataset
Optional attributes for MapDataset
Maps a function on the list of tensors unpacked from arguments on dimension 0.
 
Optional attributes for MapDefun
Op returns the number of incomplete elements in the underlying container.
 
Optional attributes for MapIncompleteSize
Op peeks at the values at the specified key.
 
Optional attributes for MapPeek
Op returns the number of elements in the underlying container.
 
Optional attributes for MapSize
Stage (key, values) in the underlying container which behaves like a hashtable.
 
Optional attributes for MapStage
Op removes and returns the values associated with the key from the underlying container.
 
Optional attributes for MapUnstage
Op removes and returns a random (key, value) from the underlying container.
 
Optional attributes for MapUnstageNoKey
Returns the set of files matching one or more glob patterns.
 
The ExperimentalMatchingFilesDataset operation
The MatchingFilesDataset operation
 
 
 
An API for building math operations as Ops
An API for building math.special operations as Ops
Multiplication matrix operations
Multiply the matrix "a" by the matrix "b".
 
Optional attributes for MatMul
Returns a batched diagonal tensor with given batched diagonal values.
 
Returns the batched diagonal part of a batched tensor.
 
Returns the batched diagonal part of a batched tensor.
 
Optional attributes for MatrixDiagPartV3
Returns a batched diagonal tensor with given batched diagonal values.
 
Optional attributes for MatrixDiagV3
Deprecated, use python implementation tf.linalg.matrix_exponential.
 
Computes the matrix logarithm of one or more square matrices: \(log(exp(A)) = A\)
 
Returns a batched matrix tensor with new batched diagonal values.
 
Optional attributes for MatrixSetDiag
Solves one or more linear least-squares problems.
 
Optional attributes for MatrixSolveLs
Computes the maximum of elements across dimensions of a tensor.
 
Optional attributes for Max
Returns the max of x and y (i.e. x > y ?
 
Creates a dataset that overrides the maximum intra-op parallelism.
Creates a dataset that overrides the maximum intra-op parallelism.
 
 
Constrains the weights incident to each hidden unit to have a norm less than or equal to a desired value.
Performs max pooling on the input.
 
Optional attributes for MaxPool
Performs 3D max pooling on the input.
 
Optional attributes for MaxPool3d
Computes gradients of 3D max pooling function.
 
Optional attributes for MaxPool3dGrad
Computes second-order gradients of the maxpooling function.
 
Optional attributes for MaxPool3dGradGrad
Computes gradients of the maxpooling function.
 
Optional attributes for MaxPoolGrad
Computes second-order gradients of the maxpooling function.
 
Optional attributes for MaxPoolGradGrad
Computes second-order gradients of the maxpooling function.
 
Optional attributes for MaxPoolGradGradWithArgmax
Computes gradients of the maxpooling function.
 
Optional attributes for MaxPoolGradWithArgmax
Performs max pooling on the input and outputs both max values and indices.
 
Optional attributes for MaxPoolWithArgmax
A metric that that implements a weighted mean MetricReduction.WEIGHTED_MEAN
Computes the mean of elements across dimensions of a tensor.
 
Optional attributes for Mean
Computes the mean of absolute difference between labels and predictions.
A metric that computes the mean of absolute difference between labels and predictions.
Computes the mean absolute percentage error between labels and predictions.
A metric that computes the mean of absolute difference between labels and predictions.
Computes the mean Intersection-Over-Union metric.
Computes the mean relative error by normalizing with the given values.
Computes the mean of squares of errors between labels and predictions.
A metric that computes the mean of absolute difference between labels and predictions.
Computes the mean squared logarithmic errors between labels and predictions.
A metric that computes the mean of absolute difference between labels and predictions.
Metric that computes the element-wise (weighted) mean of the given tensors.
 
A directory of regions in a memmapped file.
A directory of regions in a memmapped file.
A message that describes one region of memmapped file.
A message that describes one region of memmapped file.
 
 
Protobuf type tensorflow.MemoryInfo
Protobuf type tensorflow.MemoryInfo
 
Protobuf type tensorflow.MemoryLogRawAllocation
Protobuf type tensorflow.MemoryLogRawAllocation
 
Protobuf type tensorflow.MemoryLogRawDeallocation
Protobuf type tensorflow.MemoryLogRawDeallocation
 
Protobuf type tensorflow.MemoryLogStep
Protobuf type tensorflow.MemoryLogStep
 
Protobuf type tensorflow.MemoryLogTensorAllocation
Protobuf type tensorflow.MemoryLogTensorAllocation
 
Protobuf type tensorflow.MemoryLogTensorDeallocation
Protobuf type tensorflow.MemoryLogTensorDeallocation
 
Protobuf type tensorflow.MemoryLogTensorOutput
Protobuf type tensorflow.MemoryLogTensorOutput
 
For memory tracking.
For memory tracking.
 
Forwards the value of an available tensor from inputs to output.
 
An op merges elements of integer and float tensors into deduplication data as XLA tuple.
 
Optional attributes for MergeDedupData
Merges summaries.
 
V2 format specific: merges the metadata files of sharded checkpoints.
 
Optional attributes for MergeV2Checkpoints
Protocol buffer containing the following which are necessary to restart training, run inference.
Protocol buffer containing the following which are necessary to restart training, run inference.
Meta information regarding the graph to be exported.
Meta information regarding the graph to be exported.
 
 
 
Interface for metrics
Protobuf type tensorflow.MetricEntry
Protobuf type tensorflow.MetricEntry
 
Defines the different types of metric reductions
Static methods for computing metrics.
Transforms a spectrogram into a form that's useful for speech recognition.
 
Optional attributes for Mfcc
Computes the minimum of elements across dimensions of a tensor.
 
Optional attributes for Min
Returns the min of x and y (i.e. x < y ?
 
Constrains the weights to have the norm between a lower bound and an upper bound.
Pads a tensor with mirrored values.
 
Gradient op for MirrorPad op.
 
Wraps an arbitrary MLIR computation expressed as a module with a main() function.
 
Returns element-wise remainder of division.
 
 
Algorithm used for model autotuning optimization.
Protocol buffer representing the data used by the autotuning modeling framework.
Protocol buffer representing the data used by the autotuning modeling framework.
General representation of a node in the model.
General representation of a node in the model.
Represents a node parameter.
Represents a node parameter.
 
 
Contains parameters of the model autotuning optimization.
Contains parameters of the model autotuning optimization.
 
 
Class of a node in the performance model.
Identity transformation that models performance.
 
Optional attributes for ModelDataset
Stochastic gradient descent plus momentum, either nesterov or traditional.
Returns x * y element-wise.
 
Returns x * y element-wise.
 
Creates a MultiDeviceIterator resource.
 
Generates a MultiDeviceIterator resource from its provided string handle.
 
Optional attributes for MultiDeviceIteratorFromStringHandle
Gets next element for the provided shard number.
 
Initializes the multi device iterator with the given dataset.
 
Produces a string handle for the given MultiDeviceIterator.
 
Draws samples from a multinomial distribution.
 
Optional attributes for Multinomial
Creates an empty hash table that uses tensors as the backing store.
 
Optional attributes for MutableDenseHashTable
Creates an empty hash table.
 
Optional attributes for MutableHashTable
Creates an empty hash table.
 
Optional attributes for MutableHashTableOfTensors
Creates a Mutex resource that can be locked by MutexLock.
 
Optional attributes for Mutex
Locks a mutex resource.
 
Nadam Optimizer that implements the NAdam algorithm.
A list of attr names and their values.
A list of attr names and their values.
 
A pair of tensor name and tensor values.
A pair of tensor name and tensor values.
 
 
 
A Scope implementation backed by a native scope.
Deprecated.
use NcclAllReduce instead
Outputs a tensor containing the reduction across all input tensors.
 
 
Deprecated.
use NcclBroadcast instead
Sends input to all devices that are connected to the output.
 
 
Deprecated.
use NcclReduce instead
Reduces input from num_devices using reduction to a single device.
 
 
The Ndtri operation
 
Selects the k nearest centers for each point.
 
Computes numerical negative value element-wise.
 
Training via negative sampling.
 
Returns the next representable value of x1 in the direction of x2, element-wise.
 
Makes its input available to the next iteration.
 
Creates Framework nerual network Operations
An API for building nn operations as Ops
Protobuf type tensorflow.NodeDef
Protobuf type tensorflow.NodeDef
Protobuf type tensorflow.NodeDef.ExperimentalDebugInfo
Protobuf type tensorflow.NodeDef.ExperimentalDebugInfo
 
 
Time/size stats recorded for a single execution of a graph node.
Time/size stats recorded for a single execution of a graph node.
 
Output sizes recorded for a single execution of a graph node.
Output sizes recorded for a single execution of a graph node.
 
 
Non-deterministically generates some integers.
 
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.
 
Optional attributes for NonMaxSuppression
Greedily selects a subset of bounding boxes in descending order of score, pruning away boxes that have high overlaps with previously selected boxes.
 
Constrains the weights to be non-negative.
The ExperimentalNonSerializableDataset operation
The NonSerializableDataset operation
 
 
Does nothing.
 
Exception that indicates that static shapes are not able to broadcast among each other during arithmetic operations.
Returns the truth value of (x !
 
Optional attributes for NotEqual
Finds values of the n-th order statistic for the last dimension.
 
Optional attributes for NthElement
Returns a one-hot tensor.
 
Optional attributes for OneHot
Initializer that generates tensors initialized to 1.
An operator creating a constant initialized with ones of the shape given by `dims`.
Makes a "one-shot" iterator that can be iterated only once.
 
Optional attributes for OneShotIterator
Returns a tensor of ones with the same shape and type as x.
 
A logical unit of computation.
Defines an operation.
For describing inputs and outputs.
For describing inputs and outputs.
 
Description of the graph-construction-time configuration of this Op.
Description of the graph-construction-time configuration of this Op.
 
Defines an operation.
 
 
Information about version-dependent deprecation of an op
Information about version-dependent deprecation of an op
 
Interface implemented by operands of a TensorFlow operation.
Utilities for manipulating operand related types and lists.
Performs computation on Tensors.
Helper type for attribute getters, so we don't clutter the operation classes too much.
A builder for Operations.
Annotation used by classes to make TensorFlow operations conveniently accessible via org.tensorflow.op.Ops or one of its groups.
A compile-time Processor that aggregates classes annotated with org.tensorflow.op.annotation.Operator and generates the Ops convenience API.
 
An annotation to provide metadata about an op inputs accessor class.
A collection of OpDefs
A collection of OpDefs
 
An annotation to provide metadata about an op.
 
Protobuf type tensorflow.LogNormalDistribution
Protobuf type tensorflow.LogNormalDistribution
 
Protobuf type tensorflow.NormalDistribution
Protobuf type tensorflow.NormalDistribution
 
Description of an operation as well as the parameters expected to impact its performance.
Description of an operation as well as the parameters expected to impact its performance.
Input data types, shapes and values if known.
Input data types, shapes and values if known.
 
 
Performance data for tensorflow operations
Performance data for tensorflow operations
 
Memory usage data for a tensorflow operation.
Memory usage data for a tensorflow operation.
 
A collection of OpPerformance data points.
A collection of OpPerformance data points.
 
 
Description of the session when an op is run.
Description of the session when an op is run.
 
An API for building operations as Ops
A Java implementation of Scope.
Creates a dataset by applying related optimizations to input_dataset.
 
Optional attributes for OptimizeDataset
 
Optimized function graph after instantiation-related graph optimization passes (up till before graph partitioning).
Optimized function graph after instantiation-related graph optimization passes (up till before graph partitioning).
Enum for distinguishing the origin where the proto is created.
 
Base class for gradient optimizers.
A class that holds a paired gradient and variable.
Optional attributes for Optimizer
Options passed to the graph optimizer
Options passed to the graph optimizer
Control the use of the compiler/jit.
Optimization level
 
Enumerator used to create a new Optimizer with default parameters.
Constructs an Optional variant from a tuple of tensors.
 
Returns the value stored in an Optional variant or raises an error if none exists.
 
Returns true if and only if the given Optional variant has a value.
 
Creates an Optional variant with no value.
 
Creates a dataset by attaching tf.data.Options to input_dataset.
 
Optional attributes for OptionsDataset
Op removes all elements in the underlying container.
 
Optional attributes for OrderedMapClear
Op returns the number of incomplete elements in the underlying container.
 
Optional attributes for OrderedMapIncompleteSize
Op peeks at the values at the specified key.
 
Optional attributes for OrderedMapPeek
Op returns the number of elements in the underlying container.
 
Optional attributes for OrderedMapSize
Stage (key, values) in the underlying container which behaves like a ordered associative container.
 
Optional attributes for OrderedMapStage
Op removes and returns the values associated with the key from the underlying container.
 
Optional attributes for OrderedMapUnstage
Op removes and returns the (key, value) element with the smallest key from the underlying container.
 
Optional attributes for OrderedMapUnstageNoKey
A TPU core selector Op.
 
Initializer that generates an orthogonal matrix.
Retrieves a single tensor from the computation outfeed.
 
Optional attributes for OutfeedDequeue
Retrieve multiple values from the computation outfeed.
 
Optional attributes for OutfeedDequeueTuple
Retrieve multiple values from the computation outfeed.
 
Retrieves a single tensor from the computation outfeed.
 
Enqueue a Tensor on the computation outfeed.
 
Enqueue multiple Tensor values on the computation outfeed.
 
A symbolic handle to a tensor produced by an Operation.
Pads a tensor.
 
Creates a dataset that batches and pads batch_size elements from the input.
 
Optional attributes for PaddedBatchDataset
A queue that produces elements in first-in first-out order.
 
Optional attributes for PaddingFifoQueue
The ParallelBatchDataset operation
 
Optional attributes for ParallelBatchDataset
Concatenates a list of N tensors along the first dimension.
 
Interleave the values from the data tensors into a single tensor.
 
Creates a dataset containing elements of input_dataset matching predicate.
 
Optional attributes for ParallelFilterDataset
Creates a dataset that applies f to the outputs of input_dataset.
Creates a dataset that applies f to the outputs of input_dataset.
 
 
Optional attributes for ParallelInterleaveDataset
Creates a dataset that applies f to the outputs of input_dataset.
 
Optional attributes for ParallelMapDataset
Outputs random values from a normal distribution.
 
Optional attributes for ParameterizedTruncatedNormal
Transforms a vector of tf.Example protos (as strings) into typed tensors.
 
Transforms input_dataset containing Example protos as vectors of DT_STRING into a dataset of Tensor or SparseTensor objects representing the parsed features.
Transforms input_dataset containing Example protos as vectors of DT_STRING into a dataset of Tensor or SparseTensor objects representing the parsed features.
 
 
Optional attributes for ParseExampleDataset
Optional attributes for ParseExampleDataset
Transforms a vector of tf.io.SequenceExample protos (as strings) into typed tensors.
 
Optional attributes for ParseSequenceExample
Transforms a tf.Example proto (as a string) into typed tensors.
 
Transforms a scalar brain.SequenceExample proto (as strings) into typed tensors.
 
Optional attributes for ParseSingleSequenceExample
Transforms a serialized tensorflow.TensorProto proto into a Tensor.
 
returns f(inputs), where f's body is placed and partitioned.
Calls a function placed on a specified TPU device.
 
 
Optional attributes for PartitionedCall
Optional attributes for PartitionedCall
An op that groups a list of partitioned inputs together.
 
Optional attributes for PartitionedInput
An op that demultiplexes a tensor to be sharded by XLA to a list of partitioned outputs outside the XLA computation.
 
A placeholder op for a value that will be fed into the computation.
 
Optional attributes for Placeholder
A placeholder op that passes through input when its output is not fed.
 
Protobuf type tensorflow.PlatformInfo
Protobuf type tensorflow.PlatformInfo
 
Computes the Poisson loss between labels and predictions.
A metric that computes the poisson loss metric between labels and predictions.
Compute the polygamma function \(\psi^{(n)}(x)\).
 
Computes element-wise population count (a.k.a. popcount, bitsum, bitcount).
 
Computes the power of one value to another.
 
Computes the precision of the predictions with respect to the labels.
Computes best precision where recall is >= specified value.
Creates a dataset that asynchronously prefetches elements from input_dataset.
 
Optional attributes for PrefetchDataset
An op which linearizes one Tensor value to an opaque variant tensor.
 
Optional attributes for Prelinearize
An op which linearizes multiple Tensor values to an opaque variant tensor.
 
Optional attributes for PrelinearizeTuple
An identity op that triggers an error if a gradient is requested.
 
Optional attributes for PreventGradient
Prints a string scalar.
 
Optional attributes for Print
A queue that produces elements sorted by the first component value.
 
Optional attributes for PriorityQueue
Creates a dataset that uses a custom thread pool to compute input_dataset.
Creates a dataset that uses a custom thread pool to compute input_dataset.
 
 
Computes the product of elements across dimensions of a tensor.
 
Optional attributes for Prod
 
 
Next ID: 11
Next ID: 11
Protobuf enum tensorflow.ProfileOptions.DeviceType
 
Options for remote profiler session manager.
Options for remote profiler session manager.
 
Computes the QR decompositions of one or more matrices.
 
Optional attributes for Qr
An API for building quantization operations as Ops
Quantize the 'input' tensor of type float to 'output' tensor of type 'T'.
 
Optional attributes for Quantize
Quantizes then dequantizes a tensor.
 
Optional attributes for QuantizeAndDequantize
Quantizes then dequantizes a tensor.
 
Optional attributes for QuantizeAndDequantizeV3
Quantizes then dequantizes a tensor.
 
Optional attributes for QuantizeAndDequantizeV4
Returns the gradient of QuantizeAndDequantizeV4.
 
Optional attributes for QuantizeAndDequantizeV4Grad
Returns x + y element-wise, working on quantized buffers.
 
Produces the average pool of the input tensor for quantized types.
 
Quantized Batch normalization.
 
Adds Tensor 'bias' to Tensor 'input' for Quantized types.
 
Concatenates quantized tensors along one dimension.
 
Computes a 2D convolution given quantized 4D input and filter tensors.
 
Optional attributes for QuantizedConv2d
The QuantizedConv2DAndRelu operation
 
Optional attributes for QuantizedConv2DAndRelu
The QuantizedConv2DAndReluAndRequantize operation
 
Optional attributes for QuantizedConv2DAndReluAndRequantize
The QuantizedConv2DAndRequantize operation
 
Optional attributes for QuantizedConv2DAndRequantize
Computes QuantizedConv2D per channel.
 
Optional attributes for QuantizedConv2DPerChannel
The QuantizedConv2DWithBias operation
 
Optional attributes for QuantizedConv2DWithBias
The QuantizedConv2DWithBiasAndRelu operation
 
Optional attributes for QuantizedConv2DWithBiasAndRelu
The QuantizedConv2DWithBiasAndReluAndRequantize operation
 
The QuantizedConv2DWithBiasAndRequantize operation
 
Optional attributes for QuantizedConv2DWithBiasAndRequantize
The QuantizedConv2DWithBiasSignedSumAndReluAndRequantize operation
 
The QuantizedConv2DWithBiasSumAndRelu operation
 
Optional attributes for QuantizedConv2DWithBiasSumAndRelu
The QuantizedConv2DWithBiasSumAndReluAndRequantize operation
 
Computes quantized depthwise Conv2D.
 
Optional attributes for QuantizedDepthwiseConv2D
Computes quantized depthwise Conv2D with Bias.
 
Optional attributes for QuantizedDepthwiseConv2DWithBias
Computes quantized depthwise Conv2D with Bias and Relu.
 
Computes quantized depthwise Conv2D with Bias, Relu and Requantize.
 
Quantized Instance normalization.
 
Optional attributes for QuantizedInstanceNorm
Perform a quantized matrix multiplication of a by the matrix b.
 
Optional attributes for QuantizedMatMul
Performs a quantized matrix multiplication of a by the matrix b with bias add.
 
Optional attributes for QuantizedMatMulWithBias
The QuantizedMatMulWithBiasAndDequantize operation
 
Optional attributes for QuantizedMatMulWithBiasAndDequantize
Perform a quantized matrix multiplication of a by the matrix b with bias add and relu fusion.
 
Optional attributes for QuantizedMatMulWithBiasAndRelu
Perform a quantized matrix multiplication of a by the matrix b with bias add and relu and requantize fusion.
 
The QuantizedMatMulWithBiasAndRequantize operation
 
Optional attributes for QuantizedMatMulWithBiasAndRequantize
Produces the max pool of the input tensor for quantized types.
 
Returns x * y element-wise, working on quantized buffers.
 
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.
 
Computes Quantized Rectified Linear: max(features, 0)
 
Computes Quantized Rectified Linear 6: min(max(features, 0), 6)
 
Computes Quantized Rectified Linear X: min(max(features, 0), max_value)
 
Reshapes a quantized tensor as per the Reshape op.
 
Resize quantized images to size using quantized bilinear interpolation.
 
Optional attributes for QuantizedResizeBilinear
Closes the given queue.
 
Optional attributes for QueueClose
Dequeues a tuple of one or more tensors from the given queue.
 
Optional attributes for QueueDequeue
Dequeues n tuples of one or more tensors from the given queue.
 
Optional attributes for QueueDequeueMany
Dequeues n tuples of one or more tensors from the given queue.
 
Optional attributes for QueueDequeueUpTo
Enqueues a tuple of one or more tensors in the given queue.
 
Optional attributes for QueueEnqueue
Enqueues zero or more tuples of one or more tensors in the given queue.
 
Optional attributes for QueueEnqueueMany
Returns true if queue is closed.
 
Protocol buffer representing a QueueRunner.
Protocol buffer representing a QueueRunner.
 
 
Computes the number of elements in the given queue.
 
Counts the number of occurrences of each value in an integer array.
 
Optional attributes for RaggedBincount
Performs sparse-output bin counting for a ragged tensor input.
 
Optional attributes for RaggedCountSparseOutput
Generates a feature cross from a list of tensors, and returns it as a RaggedTensor.
 
The RaggedFillEmptyRows operation
 
The RaggedFillEmptyRowsGrad operation
 
Gather ragged slices from params axis 0 according to indices.
 
An API for building ragged operations as Ops
Returns a RaggedTensor containing the specified sequences of numbers.
 
Decodes a variant Tensor into a RaggedTensor.
 
Converts a RaggedTensor into a SparseTensor with the same values. input=ragged.from_nested_row_splits(rt_dense_values, rt_nested_splits) output=SparseTensor(indices=sparse_indices, values=sparse_values, dense_shape=sparse_dense_shape)
 
Create a dense tensor from a ragged tensor, possibly altering its shape.
 
Encodes a RaggedTensor into a variant Tensor.
 
Helper used to compute the gradient for RaggedTensorToVariant.
 
Randomly crop image.
 
Optional attributes for RandomCrop
Creates a Dataset that returns pseudorandom numbers.
Creates a Dataset that returns pseudorandom numbers.
 
 
Optional attributes for RandomDataset
An API for building random.experimental operations as Ops
Outputs random values from the Gamma distribution(s) described by alpha.
 
Optional attributes for RandomGamma
Computes the derivative of a Gamma random sample w.r.t.
 
Outputs the position of value in a permutation of [0, ..., max_index].
 
Optional attributes for RandomIndexShuffle
Initializer that generates tensors with a normal distribution.
An API for building random operations as Ops
Outputs random values from the Poisson distribution(s) described by rate.
 
Optional attributes for RandomPoisson
Randomly shuffles a tensor along its first dimension.
 
Optional attributes for RandomShuffle
A queue that randomizes the order of elements.
 
Optional attributes for RandomShuffleQueue
Outputs random values from a normal distribution.
 
Optional attributes for RandomStandardNormal
Initializer that generates tensors with a uniform distribution.
Outputs random values from a uniform distribution.
 
Optional attributes for RandomUniform
Outputs random integers from a uniform distribution.
 
Optional attributes for RandomUniformInt
Creates a sequence of numbers.
 
Creates a dataset with a range of values.
 
Optional attributes for RangeDataset
Returns the rank of a tensor.
 
A custom gradient for an op of unspecified type.
A base class for Op implementations that are backed by a single Operation.
A base class for operation input accessors.
A tensor which memory has not been mapped to a data space directly accessible from the JVM.
 
For serializing and restoring the state of ReaderBase, see reader_base.h for details.
For serializing and restoring the state of ReaderBase, see reader_base.h for details.
 
Returns the number of records this Reader has produced.
 
Returns the number of work units this Reader has finished processing.
 
Returns the next record (key, value pair) produced by a Reader.
 
Returns up to num_records (key, value) pairs produced by a Reader.
 
Restore a Reader to its initial clean state.
 
Restore a reader to a previously saved state.
 
Produce a string tensor that encodes the state of a Reader.
 
Reads and outputs the entire contents of the input filename.
 
Reads the value of a variable.
 
Splits resource variable input tensor across all dimensions.
 
Optional attributes for ReadVariableSplitND
Returns the real part of a complex number.
 
Returns x / y element-wise for real types.
 
Creates a dataset that changes the batch size.
 
Optional attributes for RebatchDataset
Creates a dataset that changes the batch size.
 
Computes the recall of the predictions with respect to the labels.
Computes best recall where precision is >= specified value.
Computes the reciprocal of x element-wise.
 
Computes the gradient for the inverse of x wrt its input.
 
Emits randomized records.
 
Optional attributes for RecordInput
Receives the named tensor from send_device on recv_device.
 
Optional attributes for Recv
An op that receives embedding activations on the TPU.
 
Computes the "logical and" of elements across dimensions of a tensor.
 
Optional attributes for ReduceAll
Computes the "logical or" of elements across dimensions of a tensor.
 
Optional attributes for ReduceAny
Reduces the input dataset to a singleton using a reduce function.
 
Optional attributes for ReduceDataset
Joins a string Tensor across the given dimensions.
 
Optional attributes for ReduceJoin
Reduce Log Sum Exp Operations
Computes the maximum of elements across dimensions of a tensor.
 
Optional attributes for ReduceMax
Computes the minimum of elements across dimensions of a tensor.
 
Optional attributes for ReduceMin
Computes the product of elements across dimensions of a tensor.
 
Optional attributes for ReduceProd
Computes the sum of elements across dimensions of a tensor.
 
Optional attributes for ReduceSum
Type of AbstractLoss Reduction
Creates or finds a child frame, and makes data available to the child frame.
 
Optional attributes for RefEnter
Exits the current frame to its parent frame.
 
Return the same ref tensor as the input ref tensor.
 
Forwards the value of an available tensor from inputs to output.
 
Makes its input available to the next iteration.
 
Forwards the indexth element of inputs to output.
 
Forwards the ref tensor data to the output port determined by pred.
 
Check if the input matches the regex pattern.
 
Replaces matches of the pattern regular expression in input with the replacement string provided in rewrite.
 
Optional attributes for RegexReplace
Registers a dataset with the tf.data service.
 
Optional attributes for RegisterDataset
RegisteredGradient stores a gradient function that is registered in the gradients library and used in the ops of a function in the function library.
RegisteredGradient stores a gradient function that is registered in the gradients library and used in the ops of a function in the function library.
 
 
The Relayout operation
 
The RelayoutLike operation
 
Computes rectified linear: max(features, 0).
Rectified Linear Unit(ReLU) activation.
 
Computes rectified linear 6: min(max(features, 0), 6).
 
Computes rectified linear 6 gradients for a Relu6 operation.
 
Computes rectified linear gradients for a Relu operation.
 
Runs function f on a remote device indicated by target.
 
Protobuf type tensorflow.eager.RemoteTensorHandle
Protobuf type tensorflow.eager.RemoteTensorHandle
 
 
Creates a dataset that emits the outputs of input_dataset count times.
 
Optional attributes for RepeatDataset
Connects N inputs to an N-way replicated TPU computation.
 
Optional attributes for ReplicatedInput
Connects N outputs from an N-way replicated TPU computation.
 
Metadata indicating how the TPU computation should be replicated.
 
Optional attributes for ReplicateMetadata
Computes a range that covers the actual values present in a quantized tensor.
 
Computes requantization range per channel.
 
Converts the quantized input tensor into a lower-precision output.
 
Requantizes input with min and max values known per channel.
 
Protobuf type tensorflow.RequestedExitCode
Protobuf type tensorflow.RequestedExitCode
 
Reshapes a tensor.
 
Resize images to size using area interpolation.
 
Optional attributes for ResizeArea
Resize images to size using bicubic interpolation.
 
Optional attributes for ResizeBicubic
Computes the gradient of bicubic interpolation.
 
Optional attributes for ResizeBicubicGrad
Resize images to size using bilinear interpolation.
 
Optional attributes for ResizeBilinear
Computes the gradient of bilinear interpolation.
 
Optional attributes for ResizeBilinearGrad
Resize images to size using nearest neighbor interpolation.
 
Optional attributes for ResizeNearestNeighbor
Computes the gradient of nearest neighbor interpolation.
 
Optional attributes for ResizeNearestNeighborGrad
Applies a gradient to a given accumulator.
 
Returns the number of gradients aggregated in the given accumulators.
 
Updates the accumulator with a new value for global_step.
 
Extracts the average gradient in the given ConditionalAccumulator.
 
Update '*var' according to the adadelta scheme. accum = rho() * accum + (1 - rho()) * grad.square(); update = (update_accum + epsilon).sqrt() * (accum + epsilon()).rsqrt() * grad; update_accum = rho() * update_accum + (1 - rho()) * update.square(); var -= update;
 
Optional attributes for ResourceApplyAdadelta
Update '*var' according to the adagrad scheme. accum += grad * grad var -= lr * grad * (1 / (sqrt(accum) + epsilon))
 
Optional attributes for ResourceApplyAdagrad
Update '*var' according to the proximal adagrad scheme.
 
Optional attributes for ResourceApplyAdagradDa
Update '*var' according to the Adam algorithm.
 
Optional attributes for ResourceApplyAdam
Update '*var' according to the AdaMax algorithm. m_t <- beta1 * m_{t-1} + (1 - beta1) * g v_t <- max(beta2 * v_{t-1}, abs(g)) variable <- variable - learning_rate / (1 - beta1^t) * m_t / (v_t + epsilon)
 
Optional attributes for ResourceApplyAdaMax
Update '*var' according to the Adam algorithm.
 
Optional attributes for ResourceApplyAdamWithAmsgrad
Update '*var' according to the AddSign update. m_t <- beta1 * m_{t-1} + (1 - beta1) * g update <- (alpha + sign_decay * sign(g) *sign(m)) * g variable <- variable - lr_t * update
 
Optional attributes for ResourceApplyAddSign
Update '*var' according to the centered RMSProp algorithm.
 
Optional attributes for ResourceApplyCenteredRmsProp
Update '*var' according to the Ftrl-proximal scheme. accum_new = accum + grad * grad grad_with_shrinkage = grad + 2 * l2_shrinkage * var linear += grad_with_shrinkage + (accum_new^(-lr_power) - accum^(-lr_power)) / lr * var quadratic = 1.0 / (accum_new^(lr_power) * lr) + 2 * l2 var = (sign(linear) * l1 - linear) / quadratic if |linear| > l1 else 0.0 accum = accum_new
 
Optional attributes for ResourceApplyFtrl
Update '*var' by subtracting 'alpha' * 'delta' from it.
 
Optional attributes for ResourceApplyGradientDescent
Update '*var' according to the momentum scheme.
 
Optional attributes for ResourceApplyKerasMomentum
Update '*var' according to the momentum scheme.
 
Optional attributes for ResourceApplyMomentum
Update '*var' according to the AddSign update. m_t <- beta1 * m_{t-1} + (1 - beta1) * g update <- exp(logbase * sign_decay * sign(g) * sign(m_t)) * g variable <- variable - lr_t * update
 
Optional attributes for ResourceApplyPowerSign
Update '*var' and '*accum' according to FOBOS with Adagrad learning rate. accum += grad * grad prox_v = var - lr * grad * (1 / sqrt(accum)) var = sign(prox_v)/(1+lrl2) * max{|prox_v|-lrl1,0}
 
Optional attributes for ResourceApplyProximalAdagrad
Update '*var' as FOBOS algorithm with fixed learning rate. prox_v = var - alpha * delta var = sign(prox_v)/(1+alphal2) * max{|prox_v|-alphal1,0}
 
Optional attributes for ResourceApplyProximalGradientDescent
Update '*var' according to the RMSProp algorithm.
 
Optional attributes for ResourceApplyRmsProp
A conditional accumulator for aggregating gradients.
 
Optional attributes for ResourceConditionalAccumulator
Increments variable pointed to by 'resource' until it reaches 'limit'.
 
Protobuf type tensorflow.eager.ResourceDtypeAndShape
Protobuf type tensorflow.eager.ResourceDtypeAndShape
 
Gather slices from the variable pointed to by resource according to indices.
 
Optional attributes for ResourceGather
The ResourceGatherNd operation
 
 
Protocol buffer representing a handle to a tensorflow resource.
Protocol buffer representing a handle to a tensorflow resource.
Protocol buffer representing a pair of (data type, tensor shape).
Protocol buffer representing a pair of (data type, tensor shape).
 
 
Adds sparse updates to the variable referenced by resource.
 
Divides sparse updates into the variable referenced by resource.
 
Reduces sparse updates into the variable referenced by resource using the max operation.
 
Reduces sparse updates into the variable referenced by resource using the min operation.
 
Multiplies sparse updates into the variable referenced by resource.
 
Applies sparse addition to individual values or slices in a Variable.
 
Optional attributes for ResourceScatterNdAdd
The ResourceScatterNdMax operation
 
Optional attributes for ResourceScatterNdMax
The ResourceScatterNdMin operation
 
Optional attributes for ResourceScatterNdMin
Applies sparse subtraction to individual values or slices in a Variable.
 
Optional attributes for ResourceScatterNdSub
Applies sparse updates to individual values or slices within a given variable according to indices.
 
Optional attributes for ResourceScatterNdUpdate
Subtracts sparse updates from the variable referenced by resource.
 
Assigns sparse updates to the variable referenced by resource.
 
var: Should be from a Variable().
 
Optional attributes for ResourceSparseApplyAdadelta
Update relevant entries in '*var' and '*accum' according to the adagrad scheme.
 
Optional attributes for ResourceSparseApplyAdagrad
Update entries in '*var' and '*accum' according to the proximal adagrad scheme.
 
Optional attributes for ResourceSparseApplyAdagradDa
Update relevant entries in '*var' and '*accum' according to the adagrad scheme.
 
Optional attributes for ResourceSparseApplyAdagradV2
Update '*var' according to the centered RMSProp algorithm.
 
Optional attributes for ResourceSparseApplyCenteredRmsProp
Update relevant entries in '*var' according to the Ftrl-proximal scheme.
 
Optional attributes for ResourceSparseApplyFtrl
Update relevant entries in '*var' and '*accum' according to the momentum scheme.
 
Optional attributes for ResourceSparseApplyKerasMomentum
Update relevant entries in '*var' and '*accum' according to the momentum scheme.
 
Optional attributes for ResourceSparseApplyMomentum
Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.
 
Optional attributes for ResourceSparseApplyProximalAdagrad
Sparse update '*var' as FOBOS algorithm with fixed learning rate.
 
Update '*var' according to the RMSProp algorithm.
 
Optional attributes for ResourceSparseApplyRmsProp
Assign value to the sliced l-value reference of ref.
 
Optional attributes for ResourceStridedSliceAssign
Restores tensors from a V2 checkpoint.
 
Restores a tensor from checkpoint files.
 
Optional attributes for RestoreSlice
An AutoCloseable wrapper around a Map containing Tensors.
An op that retrieves optimization parameters from embedding to host memory.
 
Retrieve Adadelta embedding parameters.
 
Retrieve Adagrad Momentum embedding parameters.
 
Retrieve Adagrad embedding parameters.
 
Retrieve ADAM embedding parameters.
 
Optional attributes for RetrieveTPUEmbeddingADAMParameters
Retrieve centered RMSProp embedding parameters.
 
Retrieve frequency estimator embedding parameters.
 
Retrieve FTRL embedding parameters.
 
Optional attributes for RetrieveTPUEmbeddingFTRLParameters
Retrieve MDL Adagrad Light embedding parameters.
 
Retrieve Momentum embedding parameters.
 
Retrieve proximal Adagrad embedding parameters.
 
The RetrieveTPUEmbeddingProximalYogiParameters operation
 
Retrieve RMSProp embedding parameters.
 
Retrieve SGD embedding parameters.
 
Reverses specific dimensions of a tensor.
 
Reverses variable length slices.
 
Optional attributes for ReverseSequence
The RewriteDataset operation
 
Graph rewriting is experimental and subject to change, not covered by any API stability guarantees.
Graph rewriting is experimental and subject to change, not covered by any API stability guarantees.
Enum for layout conversion between NCHW and NHWC on CPU.
Message to describe custom graph optimizer and its parameters
Message to describe custom graph optimizer and its parameters
 
Protobuf enum tensorflow.RewriterConfig.MemOptType
Enum controlling the number of times to run optimizers.
Protobuf enum tensorflow.RewriterConfig.Toggle
 
 
Real-valued fast Fourier transform.
 
2D real-valued fast Fourier transform.
 
3D real-valued fast Fourier transform.
 
ND fast real Fourier transform.
 
Converts one or more images from RGB to HSV.
 
Elementwise computes the bitwise right-shift of x and y.
 
Returns element-wise integer closest to x.
 
Optimizer that implements the RMSProp algorithm.
Advance the counter of a counter-based RNG.
 
Advance the counter of a counter-based RNG.
 
Rolls the elements of a tensor along an axis.
 
Computes root mean squared error metric between labels and predictions .
Rounds the values of a tensor to the nearest integer, element-wise.
 
 
 
RPC options for distributed runtime.
RPC options for distributed runtime.
 
Computes reciprocal of square root of x element-wise.
 
Computes the gradient for the rsqrt of x wrt its input.
 
Run-specific items such as arguments to the test / benchmark.
Run-specific items such as arguments to the test / benchmark.
 
Metadata output (i.e., non-Tensor) for a single Run() call.
Metadata output (i.e., non-Tensor) for a single Run() call.
Protobuf type tensorflow.RunMetadata.FunctionGraphs
Protobuf type tensorflow.RunMetadata.FunctionGraphs
 
 
Options for a single Run() call.
Options for a single Run() call.
Everything inside Experimental is subject to change and is not subject to API stability guarantees in https://www.tensorflow.org/guide/version_compat.
Everything inside Experimental is subject to change and is not subject to API stability guarantees in https://www.tensorflow.org/guide/version_compat.
Options for run handler thread pool.
Options for run handler thread pool.
 
 
TODO(pbar) Turn this into a TraceOptions proto which allows tracing to be controlled in a more orthogonal manner?
 
Generate a single randomly distorted bounding box for an image.
 
Optional attributes for SampleDistortedBoundingBox
Creates a dataset that takes a Bernoulli sample of the contents of another dataset.
 
Saves tensors in V2 checkpoint format.
 
The SaveDatasetV2 operation
 
Optional attributes for SaveDataset
SavedModel is the high level serialization format for TensorFlow Models.
SavedModel is the high level serialization format for TensorFlow Models.
SavedModelBundle represents a model loaded from storage.
Options for exporting a SavedModel.
Options for loading a SavedModel.
 
 
 
Protobuf type tensorflow.CapturedTensor
Protobuf type tensorflow.CapturedTensor
 
Represents `FunctionSpec` used in `Function`.
Represents `FunctionSpec` used in `Function`.
Whether the function should be compiled by XLA.
 
Protobuf type tensorflow.SaveableObject
Protobuf type tensorflow.SaveableObject
 
A SavedAsset points to an asset in the MetaGraph.
A SavedAsset points to an asset in the MetaGraph.
 
Protobuf type tensorflow.SavedBareConcreteFunction
Protobuf type tensorflow.SavedBareConcreteFunction
 
Stores low-level information about a concrete function.
Stores low-level information about a concrete function.
 
Protobuf type tensorflow.SavedConstant
Protobuf type tensorflow.SavedConstant
 
A function with multiple signatures, possibly with non-Tensor arguments.
A function with multiple signatures, possibly with non-Tensor arguments.
 
Protobuf type tensorflow.SavedObject
Protobuf type tensorflow.SavedObject
 
Protobuf type tensorflow.SavedObjectGraph
Protobuf type tensorflow.SavedObjectGraph
 
 
A SavedResource represents a TF object that holds state during its lifetime.
A SavedResource represents a TF object that holds state during its lifetime.
 
A SavedUserObject is an object (in the object-oriented language of the TensorFlow program) of some user- or framework-defined class other than those handled specifically by the other kinds of SavedObjects.
A SavedUserObject is an object (in the object-oriented language of the TensorFlow program) of some user- or framework-defined class other than those handled specifically by the other kinds of SavedObjects.
 
Represents a Variable that is initialized by loading the contents from the checkpoint.
Represents a Variable that is initialized by loading the contents from the checkpoint.
 
Saved tensor slice: it stores the name of the tensors, the slice, and the raw data.
Saved tensor slice: it stores the name of the tensors, the slice, and the raw data.
Metadata describing the set of slices of the same tensor saved in a checkpoint file.
Metadata describing the set of slices of the same tensor saved in a checkpoint file.
 
 
Metadata describing the set of tensor slices saved in a checkpoint file.
Metadata describing the set of tensor slices saved in a checkpoint file.
 
 
Each record in a v3 checkpoint file is a serialized SavedTensorSlices message.
Each record in a v3 checkpoint file is a serialized SavedTensorSlices message.
 
Protocol buffer representing the configuration of a Saver.
Protocol buffer representing the configuration of a Saver.
A version number that identifies a different on-disk checkpoint format.
 
 
Protobuf type tensorflow.SaveSliceInfoDef
Protobuf type tensorflow.SaveSliceInfoDef
 
Saves input tensors slices to disk.
 
Outputs a Summary protocol buffer with scalar values.
 
The ScaleAndTranslate operation
 
Optional attributes for ScaleAndTranslate
The ScaleAndTranslateGrad operation
 
Optional attributes for ScaleAndTranslateGrad
Creates a dataset successively reduces f over the elements of input_dataset.
Creates a dataset successively reduces f over the elements of input_dataset.
 
 
Optional attributes for ScanDataset
Optional attributes for ScanDataset
Adds sparse updates to a variable reference.
 
Optional attributes for ScatterAdd
Divides a variable reference by sparse updates.
 
Optional attributes for ScatterDiv
Reduces sparse updates into a variable reference using the max operation.
 
Optional attributes for ScatterMax
Reduces sparse updates into a variable reference using the min operation.
 
Optional attributes for ScatterMin
Multiplies sparse updates into a variable reference.
 
Optional attributes for ScatterMul
Scatters updates into a tensor of shape shape according to indices.
 
Optional attributes for ScatterNd
Applies sparse addition to individual values or slices in a Variable.
 
Optional attributes for ScatterNdAdd
Computes element-wise maximum.
 
Optional attributes for ScatterNdMax
Computes element-wise minimum.
 
Optional attributes for ScatterNdMin
Applies sparse addition to input using individual values or slices from updates according to indices indices.
 
Optional attributes for ScatterNdNonAliasingAdd
Applies sparse subtraction to individual values or slices in a Variable. within a given variable according to indices.
 
Optional attributes for ScatterNdSub
Applies sparse updates to individual values or slices within a given variable according to indices.
 
Optional attributes for ScatterNdUpdate
Subtracts sparse updates to a variable reference.
 
Optional attributes for ScatterSub
Applies sparse updates to a variable reference.
 
Optional attributes for ScatterUpdate
Manages groups of related properties when creating Tensorflow Operations, such as a common name prefix.
Protobuf type tensorflow.ScopedAllocatorOptions
Protobuf type tensorflow.ScopedAllocatorOptions
 
Computes fingerprints of the input strings.
 
Distributed version of Stochastic Dual Coordinate Ascent (SDCA) optimizer for linear models with L1 + L2 regularization.
 
Optional attributes for SdcaOptimizer
Applies L1 regularization shrink step on the parameters.
 
Computes the maximum along segments of a tensor.
 
Computes the mean along segments of a tensor.
 
Computes the minimum along segments of a tensor.
 
Computes the product along segments of a tensor.
 
Computes the sum along segments of a tensor.
 
The SelectV2 operation
 
Computes the eigen decomposition of one or more square self-adjoint matrices.
 
Optional attributes for SelfAdjointEig
Computes scaled exponential linear: scale * alpha * (exp(features) - 1) if < 0, scale * features otherwise.
Scaled Exponential Linear Unit (SELU).
 
Computes gradients for the scaled exponential linear (Selu) operation.
 
Sends the named tensor from send_device to recv_device.
 
Optional attributes for Send
Performs gradient updates of embedding tables.
 
Optional attributes for SendTPUEmbeddingGradients
Computes best sensitivity where sensitivity is >= specified value.
Protobuf type tensorflow.SequenceExample
Protobuf type tensorflow.SequenceExample
 
Represents a serialized tf.dtypes.Dtype
Represents a serialized tf.dtypes.Dtype
 
Converts the given resource_handle representing an iterator to a variant tensor.
 
Optional attributes for SerializeIterator
Serialize an N-minibatch SparseTensor into an [N, 3] Tensor object.
 
Serialize a SparseTensor into a [3] Tensor object.
 
Transforms a Tensor into a serialized TensorProto proto.
 
An in-process TensorFlow server, for use in distributed training.
Defines the configuration of a single TensorFlow server.
Defines the configuration of a single TensorFlow server.
 
 
 
Configuration for a tf.data service DispatchServer.
Configuration for a tf.data service DispatchServer.
 
Configuration for a tf.data service WorkerServer.
Configuration for a tf.data service WorkerServer.
 
Driver for Graph execution.
A callable function backed by a session.
Protocol buffer used for logging session state.
Protocol buffer used for logging session state.
Protobuf enum tensorflow.SessionLog.SessionStatus
 
Metadata about the session.
Metadata about the session.
 
Computes the difference between two lists of numbers or strings.
 
Creates Framework set Operations
 
Enumeration containing the string operation values to be passed to the TensorFlow Sparse Ops function SparseOps.denseToDenseSetOperation(Operand, Operand, String, DenseToDenseSetOperation.Options...)
Number of unique elements along last dimension of input set.
 
Optional attributes for SetSize
The ExperimentalSetStatsAggregatorDataset operation
The SetStatsAggregatorDataset operation
 
 
Returns the shape of a tensor.
 
Returns shape of tensors.
 
An API for building shape operations as Ops
An operator providing methods on org.tensorflow.op.core.Shape tensors and 1d operands that represent the dimensions of a shape.
Various methods for processing with Shapes and Operands
Creates a Dataset that includes only 1/num_shards of this dataset.
 
Optional attributes for ShardDataset
Generate a sharded filename.
 
Generate a glob pattern matching all sharded file names.
 
The ShuffleAndRepeatDatasetV2 operation
 
Optional attributes for ShuffleAndRepeatDataset
The ShuffleDatasetV3 operation
 
Optional attributes for ShuffleDataset
Shuts down a running distributed TPU system.
 
An op that shuts down the TPU system.
 
Sigmoid activation.
Computes sigmoid of x element-wise.
 
 
Computes the gradient of the sigmoid of x wrt its input.
 
Returns an element-wise indication of the sign of a number.
 
An API for building signal operations as Ops
Describe the inputs and outputs of an executable entity, such as a ConcreteFunction, among other useful metadata.
Builds a new function signature.
 
SignatureDef defines the signature of a computation supported by a TensorFlow graph.
SignatureDef defines the signature of a computation supported by a TensorFlow graph.
 
Computes sine of x element-wise.
 
Computes hyperbolic sine of x element-wise.
 
Returns the size of a tensor.
 
Creates a dataset that skips count elements from the input_dataset.
 
Optional attributes for SkipDataset
Parses a text file and creates a batch of examples.
 
Optional attributes for Skipgram
The ExperimentalSleepDataset operation
The SleepDataset operation
 
 
Return a slice from 'input'.
 
Creates a dataset that passes a sliding window over input_dataset.
Creates a dataset that passes a sliding window over input_dataset.
 
 
Optional attributes for SlidingWindowDataset
Returns a copy of the input tensor.
 
Metadata for a `tf.data.Dataset` distributed snapshot.
Metadata for a `tf.data.Dataset` distributed snapshot.
 
 
This stores the metadata information present in each snapshot record.
This stores the metadata information present in each snapshot record.
 
Each SnapshotRecord represents one batch of pre-processed input data.
Each SnapshotRecord represents one batch of pre-processed input data.
 
Metadata for all the tensors in a Snapshot Record.
Metadata for all the tensors in a Snapshot Record.
 
Metadata for a single tensor in the Snapshot Record.
Metadata for a single tensor in the Snapshot Record.
 
The SnapshotChunkDataset operation
 
Optional attributes for SnapshotChunkDataset
Creates a dataset that will write to / read from a snapshot.
 
Optional attributes for SnapshotDataset
The SnapshotDatasetReader operation
 
Optional attributes for SnapshotDatasetReader
The SnapshotNestedDatasetReader operation
 
Generates points from the Sobol sequence.
 
Softmax converts a real vector to a vector of categorical probabilities.
Computes softmax activations.
 
 
Computes softmax cross entropy cost and gradients to backpropagate.
 
Softplus activation function, softplus(x) = log(exp(x) + 1).
The Softplus operation
 
Computes softplus gradients for a softplus operation.
 
Softsign activation function, softsign(x) = x / (abs(x) + 1).
Computes softsign: features / (abs(features) + 1).
 
Computes softsign gradients for a softsign operation.
 
Solves systems of linear equations.
 
Optional attributes for Solve
The SortListOfSparseCoreCooTensors operation
 
Content of a source file involved in the execution of the debugged TensorFlow program.
Content of a source file involved in the execution of the debugged TensorFlow program.
 
Holds the information of the source that writes the events.
Holds the information of the source that writes the events.
 
SpaceToBatch for 4-D tensors of type T.
 
SpaceToBatch for N-D tensors of type T.
 
SpaceToDepth for tensors of type T.
 
Optional attributes for SpaceToDepth
Applies a sparse gradient to a given accumulator.
 
Extracts the average sparse gradient in a SparseConditionalAccumulator.
 
Adds two SparseTensor objects to produce another SparseTensor.
 
The gradient operator for the SparseAdd op.
 
var: Should be from a Variable().
 
Optional attributes for SparseApplyAdadelta
Update relevant entries in '*var' and '*accum' according to the adagrad scheme.
 
Optional attributes for SparseApplyAdagrad
Update entries in '*var' and '*accum' according to the proximal adagrad scheme.
 
Optional attributes for SparseApplyAdagradDa
Update '*var' according to the centered RMSProp algorithm.
 
Optional attributes for SparseApplyCenteredRmsProp
Update relevant entries in '*var' according to the Ftrl-proximal scheme.
 
Optional attributes for SparseApplyFtrl
Update relevant entries in '*var' and '*accum' according to the momentum scheme.
 
Optional attributes for SparseApplyMomentum
Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.
 
Optional attributes for SparseApplyProximalAdagrad
Sparse update '*var' as FOBOS algorithm with fixed learning rate.
 
Optional attributes for SparseApplyProximalGradientDescent
Update '*var' according to the RMSProp algorithm.
 
Optional attributes for SparseApplyRmsProp
Counts the number of occurrences of each value in an integer array.
 
Optional attributes for SparseBincount
Calculates how often predictions matches integer labels.
Computes the crossentropy loss between labels and predictions.
A metric that computes the sparse categorical cross-entropy loss between true labels and predicted labels.
Concatenates a list of SparseTensor along the specified dimension.
 
A conditional accumulator for aggregating sparse gradients.
 
Optional attributes for SparseConditionalAccumulator
Performs sparse-output bin counting for a sparse tensor input.
 
Optional attributes for SparseCountSparseOutput
Generates sparse cross from a list of sparse and dense tensors.
 
Generates sparse cross from a list of sparse and dense tensors.
 
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.
 
Component-wise divides a SparseTensor by a dense Tensor.
 
Component-wise multiplies a SparseTensor by a dense Tensor.
 
Fills empty rows in the input 2-D SparseTensor with a default value.
 
The gradient of SparseFillEmptyRows.
 
Multiply matrix "a" by matrix "b".
 
Optional attributes for SparseMatMul
Sparse addition of two CSR matrices, C = alpha * A + beta * B.
 
Matrix-multiplies a sparse matrix with a dense matrix.
 
Optional attributes for SparseMatrixMatMul
Element-wise multiplication of a sparse matrix with a dense tensor.
 
Returns the number of nonzeroes of sparse_matrix.
 
Computes the Approximate Minimum Degree (AMD) ordering of input.
 
Calculates the softmax of a CSRSparseMatrix.
 
Calculates the gradient of the SparseMatrixSoftmax op.
 
Computes the sparse Cholesky decomposition of input.
 
Sparse-matrix-multiplies two CSR matrices a and b.
 
Optional attributes for SparseMatrixSparseMatMul
Transposes the inner (matrix) dimensions of a CSRSparseMatrix.
 
Optional attributes for SparseMatrixTranspose
Creates an all-zeros CSRSparseMatrix with shape dense_shape.
 
An API for building sparse operations as Ops
Computes the max of elements across dimensions of a SparseTensor.
 
Optional attributes for SparseReduceMax
Computes the max of elements across dimensions of a SparseTensor.
 
Optional attributes for SparseReduceMaxSparse
Computes the sum of elements across dimensions of a SparseTensor.
 
Optional attributes for SparseReduceSum
Computes the sum of elements across dimensions of a SparseTensor.
 
Optional attributes for SparseReduceSumSparse
Reorders a SparseTensor into the canonical, row-major ordering.
 
Reshapes a SparseTensor to represent values in a new dense shape.
 
Computes the mean along sparse segments of a tensor.
 
Optional attributes for SparseSegmentMean
Computes gradients for SparseSegmentMean.
 
Computes the mean along sparse segments of a tensor.
 
Optional attributes for SparseSegmentMeanWithNumSegments
Computes the sum along sparse segments of a tensor divided by the sqrt of N.
 
Optional attributes for SparseSegmentSqrtN
Computes gradients for SparseSegmentSqrtN.
 
Computes the sum along sparse segments of a tensor divided by the sqrt of N.
 
Optional attributes for SparseSegmentSqrtNWithNumSegments
Computes the sum along sparse segments of a tensor.
 
Optional attributes for SparseSegmentSum
Computes gradients for SparseSegmentSum.
 
Computes the sum along sparse segments of a tensor.
 
Optional attributes for SparseSegmentSumWithNumSegments
Slice a SparseTensor based on the start and size.
 
The gradient operator for the SparseSlice op.
 
Applies softmax to a batched N-D SparseTensor.
 
 
Computes softmax cross entropy cost and gradients to backpropagate.
 
Returns the element-wise max of two SparseTensors.
 
Returns the element-wise min of two SparseTensors.
 
Split a SparseTensor into num_split tensors along one dimension.
 
This is a helper class that represents a sparse tensor who's attributes may be passed to SparseOps methods.
A virtual type of Tensor composed of three dense tensors (indices, values and dimensions) used to represent the sparse data into a multi-dimensional dense space.
Adds up a SparseTensor and a dense Tensor, producing a dense Tensor.
 
Multiply SparseTensor (of rank 2) "A" by dense matrix "B".
 
Optional attributes for SparseTensorDenseMatMul
Creates a dataset that splits a SparseTensor into elements row-wise.
 
Converts a SparseTensor to a (possibly batched) CSRSparseMatrix.
 
Converts a sparse representation into a dense tensor.
 
Optional attributes for SparseToDense
Computes how often integer targets are in the top `K` predictions.
Applies set operation along last dimension of 2 SparseTensor inputs.
 
Optional attributes for SparseToSparseSetOperation
Computes best specificity where sensitivity is >= specified value.
The Spence operation
 
Splits a tensor into num_split tensors along one dimension.
 
An op splits input deduplication data XLA tuple into integer and floating point tensors.
 
Optional attributes for SplitDedupData
Splits input tensor across all dimensions.
 
Optional attributes for SplitND
Splits a tensor into num_split tensors along one dimension.
 
Creates a dataset that executes a SQL query and emits rows of the result set.
Creates a dataset that executes a SQL query and emits rows of the result set.
 
 
Computes square root of x element-wise.
 
Computes the gradient for the sqrt of x wrt its input.
 
Computes the matrix square root of one or more square matrices: matmul(sqrtm(A), sqrtm(A)) = A
 
Computes square of x element-wise.
 
Returns conj(x - y)(x - y) element-wise.
 
Computes the squared hinge loss between labels and predictions.
A metric that computes the squared hinge loss metric between labels and predictions.
Removes dimensions of size 1 from the shape of a tensor.
 
Optional attributes for Squeeze
Packs a list of N rank-R tensors into one rank-(R+1) tensor.
 
Optional attributes for Stack
Delete the stack from its resource container.
 
A stack that produces elements in first-in last-out order.
 
Optional attributes for StackCreate
A stack frame with ID.
A stack frame with ID.
 
Pop the element at the top of the stack.
 
Push an element onto the stack.
 
Optional attributes for StackPush
Stage values similar to a lightweight Enqueue.
 
Optional attributes for Stage
Op removes all elements in the underlying container.
 
Optional attributes for StageClear
Op peeks at the values at the specified index.
 
Optional attributes for StagePeek
Op returns the number of elements in the underlying container.
 
Optional attributes for StageSize
An n-way switch statement which calls a single branch function.
 
output = cond ?
 
returns f(inputs), where f's body is placed and partitioned.
 
Optional attributes for StatefulPartitionedCall
The StatefulRandomBinomial operation
 
Outputs random values from a normal distribution.
 
Outputs random values from a truncated normal distribution.
 
Outputs random values from a uniform distribution.
 
Outputs random integers from a uniform distribution.
 
Outputs random integers from a uniform distribution.
 
output = input; While (Cond(output)) { output = Body(output) }
 
An n-way switch statement which calls a single branch function.
 
output = cond ?
 
Draws samples from a multinomial distribution.
 
The StatelessParameterizedTruncatedNormal operation
 
Outputs deterministic pseudorandom random numbers from a binomial distribution.
 
Outputs deterministic pseudorandom random numbers from a gamma distribution.
 
Picks the best counter-based RNG algorithm based on device.
 
Scrambles seed into key and counter, using the best algorithm based on device.
 
Picks the best algorithm based on device, and scrambles seed into key and counter.
 
Outputs deterministic pseudorandom values from a normal distribution.
 
Outputs deterministic pseudorandom values from a normal distribution.
 
Outputs deterministic pseudorandom random numbers from a Poisson distribution.
 
Outputs deterministic pseudorandom random values from a uniform distribution.
 
Outputs deterministic pseudorandom random integers from a uniform distribution.
 
Outputs deterministic pseudorandom random integers from a uniform distribution.
 
Outputs deterministic pseudorandom random integers from a uniform distribution.
 
Outputs deterministic pseudorandom random integers from a uniform distribution.
 
Outputs deterministic pseudorandom random values from a uniform distribution.
 
Generate a randomly distorted bounding box for an image deterministically.
 
Optional attributes for StatelessSampleDistortedBoundingBox
Randomly and deterministically shuffles a tensor along its first dimension.
 
Outputs deterministic pseudorandom values from a truncated normal distribution.
 
Outputs deterministic pseudorandom values from a truncated normal distribution.
 
output = input; While (Cond(output)) { output = Body(output) }
 
Check if the input matches the regex pattern.
 
Replaces the match of pattern in input with rewrite.
 
Optional attributes for StaticRegexReplace
Creates a statistics manager resource.
The StatsAggregatorHandleV2 operation
 
 
Optional attributes for StatsAggregatorHandle
Optional attributes for StatsAggregatorHandle
Set a summary_writer_interface to record statistics using given stats_aggregator.
 
Produces a summary of any statistics recorded by the given statistics manager.
Produces a summary of any statistics recorded by the given statistics manager.
 
 
 
 
Wire-format for Status.
Wire-format for Status.
 
Protobuf type tensorflow.StepStats
Protobuf type tensorflow.StepStats
 
 
Stochastically cast a given tensor from floats to ints.
 
Stops gradient computation.
 
The StoreMinibatchStatisticsInFdo operation
 
Return a strided slice from input.
 
Optional attributes for StridedSlice
Assign value to the sliced l-value reference of ref.
 
Optional attributes for StridedSliceAssign
Returns the gradient of StridedSlice.
 
Optional attributes for StridedSliceGrad
Helper endpoint methods for Python like indexing.
Formats a string template using a list of tensors.
 
Optional attributes for StringFormat
String lengths of input.
 
Optional attributes for StringLength
Creates ngrams from ragged string data.
 
An API for building strings operations as Ops
Split elements of source based on sep into a SparseTensor.
 
Optional attributes for StringSplit
Strip leading and trailing whitespaces from the Tensor.
 
 
A protobuf to represent tf.BoundedTensorSpec.
A protobuf to represent tf.BoundedTensorSpec.
 
Represents a Python dict keyed by `str`.
Represents a Python dict keyed by `str`.
 
Represents a Python list.
Represents a Python list.
 
Represents Python's namedtuple.
Represents Python's namedtuple.
 
Represents None.
Represents None.
 
Represents a (key, value) pair.
Represents a (key, value) pair.
 
`StructuredValue` represents a dynamically typed value representing various data structures that are inspired by Python data structures typically used in TensorFlow functions as inputs and outputs.
`StructuredValue` represents a dynamically typed value representing various data structures that are inspired by Python data structures typically used in TensorFlow functions as inputs and outputs.
 
 
A protobuf to represent tf.TensorSpec.
A protobuf to represent tf.TensorSpec.
 
Represents a Python tuple.
Represents a Python tuple.
 
Represents a tf.TypeSpec
Represents a tf.TypeSpec
Protobuf enum tensorflow.TypeSpecProto.TypeSpecClass
 
Returns x - y element-wise.
 
Return substrings from Tensor of strings.
 
Optional attributes for Substr
Computes the (weighted) sum of the given values.
Computes the sum of elements across dimensions of a tensor.
 
Optional attributes for Sum
A Summary is a set of named values to be displayed by the visualizer.
Protobuf type tensorflow.Summary.Audio
Protobuf type tensorflow.Summary.Audio
 
A Summary is a set of named values to be displayed by the visualizer.
Protobuf type tensorflow.Summary.Image
Protobuf type tensorflow.Summary.Image
 
Protobuf type tensorflow.Summary.Value
Protobuf type tensorflow.Summary.Value
 
 
Metadata associated with a series of Summary data
Metadata associated with a series of Summary data
 
A SummaryMetadata encapsulates information on which plugins are able to make use of a certain summary value.
A SummaryMetadata encapsulates information on which plugins are able to make use of a certain summary value.
Protobuf type tensorflow.SummaryMetadata.PluginData
Protobuf type tensorflow.SummaryMetadata.PluginData
 
 
An API for building summary operations as Ops
 
 
The SummaryWriter operation
 
Optional attributes for SummaryWriter
Computes the singular value decompositions of one or more matrices.
 
Optional attributes for Svd
Swish activation function.
Forwards data to the output port determined by pred.
 
Computes the gradient function for function f via backpropagation.
 
Synchronizes the device this op is run on.
 
For logging the metadata output for a single session.run() call.
For logging the metadata output for a single session.run() call.
 
Creates a dataset that contains count elements from the input_dataset.
 
Optional attributes for TakeDataset
Read SparseTensors from a SparseTensorsMap and concatenate them.
 
Optional attributes for TakeManySparseFromTensorsMap
Creates a dataset that stops iteration when predicate` is false.
Creates a dataset that stops iteration when predicate` is false.
 
 
Optional attributes for TakeWhileDataset
Computes tan of x element-wise.
 
Hyperbolic tangent activation function.
Computes hyperbolic tangent of x element-wise.
 
Computes the gradient for the tanh of x wrt its input.
 
Defines the device filters for a remote task.
Defines the device filters for a remote task.
 
Brain 16-bit float tensor type.
Boolean tensor type.
Returns a tensor that may be mutated, but only persists within a single step.
 
Optional attributes for TemporaryVariable
 
A statically typed multi-dimensional array.
An array of Tensors of given size.
 
Optional attributes for TensorArray
Delete the TensorArray from its resource container.
 
Concat the elements from the TensorArray into value value.
 
Optional attributes for TensorArrayConcat
Gather specific elements from the TensorArray into output value.
 
Optional attributes for TensorArrayGather
Creates a TensorArray for storing the gradients of values in the given handle.
 
Creates a TensorArray for storing multiple gradients of values in the given handle.
 
The TensorArrayPack operation
 
Optional attributes for TensorArrayPack
Read an element from the TensorArray into output value.
 
Scatter the data from the input value into specific TensorArray elements.
 
Get the current size of the TensorArray.
 
Split the data from the input value into TensorArray elements.
 
The TensorArrayUnpack operation
 
Push an element onto the tensor_array.
 
 
Defines a connection between two tensors in a `GraphDef`.
Defines a connection between two tensors in a `GraphDef`.
 
Creates a dataset that emits components as a tuple of tensors once.
 
Optional attributes for TensorDataset
Available modes for extracting debugging information from a Tensor.
Protobuf type tensorflow.TensorDescription
Protobuf type tensorflow.TensorDescription
 
 
Returns a diagonal tensor with a given diagonal values.
 
Returns the diagonal part of the tensor.
 
tensor contraction Operations
 
 
Static utility methods describing the TensorFlow runtime.
Unchecked exception thrown by TensorFlow core classes
A function that can be called with tensors.
Information about a Tensor necessary for feeding or retrieval.
Information about a Tensor necessary for feeding or retrieval.
Generic encoding for composite tensors.
Generic encoding for composite tensors.
 
For sparse tensors, The COO encoding stores a triple of values, indices, and shape.
For sparse tensors, The COO encoding stores a triple of values, indices, and shape.
 
 
 
Concats all tensors in the list along the 0th dimension.
 
The TensorListConcatLists operation
 
The shape of the elements of the given list, as a tensor. input_handle: the list element_shape: the shape of elements of the list
 
Creates a TensorList which, when stacked, has the value of tensor.
 
Creates a Tensor by indexing into the TensorList.
 
Returns the item in the list with the given index. input_handle: the list index: the position in the list from which an element will be retrieved item: the element at that position
 
Returns the number of tensors in the input tensor list. input_handle: the input list length: the number of tensors in the list
 
Returns the last element of the input list as well as a list with all but that element.
 
Returns a list which has the passed-in Tensor as last element and the other elements of the given list in input_handle. tensor: The tensor to put on the list. input_handle: The old list. output_handle: A list with the elements of the old list followed by tensor. element_dtype: the type of elements in the list. element_shape: a shape compatible with that of elements in the list.
 
The TensorListPushBackBatch operation
 
List of the given size with empty elements. element_shape: the shape of the future elements of the list num_elements: the number of elements to reserve handle: the output list element_dtype: the desired type of elements in the list.
 
Resizes the list. input_handle: the input list size: size of the output list
 
Creates a TensorList by indexing into a Tensor.
 
Scatters tensor at indices in an input list.
 
Sets the index-th position of the list to contain the given tensor. input_handle: the list index: the position in the list to which the tensor will be assigned item: the element to be assigned to that position output_handle: the new list, with the element in the proper position
 
Optional attributes for TensorListSetItem
Splits a tensor into a list. list[i] corresponds to lengths[i] tensors from the input tensor.
 
Stacks all tensors in the list.
 
Optional attributes for TensorListStack
Returns a tensor map with item from given key erased. input_handle: the original map output_handle: the map with value from given key removed key: the key of the value to be erased
 
Returns whether the given key exists in the map. input_handle: the input map key: the key to check has_key: whether the key is already in the map or not
 
Returns a map that is the 'input_handle' with the given key-value pair inserted. input_handle: the original map output_handle: the map with key and value inserted key: the key to be inserted value: the value to be inserted
 
Returns the value from a given key in a tensor map. input_handle: the input map key: the key to be looked up value: the value found from the given key
 
Maps the native memory of a RawTensor to a n-dimensional typed data space accessible from the JVM.
Returns the number of tensors in the input tensor map. input_handle: the input map size: the number of tensors in the map
 
Returns a Tensor stack of all keys in a tensor map. input_handle: the input map keys: the returned Tensor of all keys in the map
 
Protocol buffer representing a tensor.
Protocol buffer representing a tensor.
 
 
Adds sparse updates to an existing tensor according to indices.
 
Optional attributes for TensorScatterNdAdd
Apply a sparse update to a tensor taking the element-wise maximum.
 
Optional attributes for TensorScatterNdMax
The TensorScatterMin operation
 
Optional attributes for TensorScatterNdMin
Subtracts sparse updates from an existing tensor according to indices.
 
Optional attributes for TensorScatterNdSub
Scatter updates into an existing tensor according to indices.
 
Optional attributes for TensorScatterNdUpdate
Dimensions of a tensor.
Dimensions of a tensor.
One dimension of the tensor.
One dimension of the tensor.
 
 
 
Creates a dataset that emits each dim-0 slice of components once.
 
Optional attributes for TensorSliceDataset
Can only be interpreted if you know the corresponding TensorShape.
Can only be interpreted if you know the corresponding TensorShape.
Extent of the slice in one dimension.
Extent of the slice in one dimension.
 
 
 
 
Assign value to the sliced l-value reference of input.
 
Optional attributes for TensorStridedSliceUpdate
Outputs a Summary protocol buffer with a tensor and per-plugin data.
 
Annotation for all tensor types.
 
 
The output of one benchmark / test run.
The type of benchmark.
The output of one benchmark / test run.
 
Creates a dataset that emits the lines of one or more text files.
 
Optional attributes for TextLineDataset
A Reader that outputs the lines of a file delimited by '\n'.
 
Optional attributes for TextLineReader
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Function to be implemented on the JVM side to be called back by the native library when it is time to attach gradient operations for the given op, graph and scope.
Unique identifier of a TensorFlow graph instance
Helper class for loading the TensorFlow Java native library.
 
IEEE-754 half-precision 16-bit float tensor type.
IEEE-754 single-precision 32-bit float tensor type.
IEEE-754 double-precision 64-bit float tensor type.
Common interface for all floating point tensors.
 
 
Creates a dataset that emits the records from one or more TFRecord files.
 
Optional attributes for TfRecordDataset
A Reader that outputs the records from a TensorFlow Records file.
 
Optional attributes for TfRecordReader
 
 
 
Creates a dataset that uses a custom thread pool to compute input_dataset.
Creates a dataset that uses a custom thread pool to compute input_dataset.
 
 
Creates a dataset that uses a custom thread pool to compute input_dataset.
Creates a dataset that uses a custom thread pool to compute input_dataset.
 
 
Optional attributes for ThreadPoolHandle
Optional attributes for ThreadPoolHandle
Protobuf type tensorflow.ThreadPoolOptionProto
Protobuf type tensorflow.ThreadPoolOptionProto
 
Generates labels for candidate sampling with a learned unigram distribution.
 
Optional attributes for ThreadUnsafeUnigramCandidateSampler
Constructs a tensor by tiling a given tensor.
 
Returns the gradient of Tile.
 
Provides the time since epoch in seconds.
 
32-bit signed integer tensor type.
64-bit signed integer tensor type.
Common interface for all integral numeric tensors.
Common interface for all numeric tensors.
Converts a tensor to a scalar predicate.
 
Converts each string in the input Tensor to its hash mod by a number of buckets.
 
Converts each string in the input Tensor to its hash mod by a number of buckets.
 
Converts each string in the input Tensor to its hash mod by a number of buckets.
 
Converts each string in the input Tensor to the specified numeric type.
 
Finds values and indices of the k largest elements for the last dimension.
 
Optional attributes for TopK
Computes the poisson loss metric between labels and predictions.
Returns the TopK unique values in the array in sorted order.
 
Returns the TopK values in the array in sorted order.
 
The TPUAnnotateTensorsWithDynamicShape operation
 
Deprecated.
use CompilationResult instead
 
Op that copies host tensor to device with dynamic shape support.
 
Deprecated.
 
Converts XRT's uid handles to TensorFlow-friendly input format.
 
An API for building tpu operations as Ops
Deprecated.
use ReplicatedInput instead
 
Optional attributes for TPUReplicatedInput
Deprecated.
use ReplicatedOutput instead
 
Deprecated.
use ReplicateMetadata instead
 
Optional attributes for TPUReplicateMetadata
Op that reshards on-device TPU variables to specified state.
 
Round-robin load balancing on TPU cores.
 
 
Protobuf type tensorflow.RegisteredSaver
Protobuf type tensorflow.RegisteredSaver
 
Protobuf type tensorflow.TrackableObjectGraph
Protobuf type tensorflow.TrackableObjectGraph
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject.ObjectReference
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject.ObjectReference
 
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject.SerializedTensor
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject.SerializedTensor
 
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject.SlotVariableReference
Protobuf type tensorflow.TrackableObjectGraph.TrackableObject.SlotVariableReference
 
 
 
An API for building train operations as Ops
 
Extra data needed on a non-RDMA RecvBufResponse.
Extra data needed on a non-RDMA RecvBufResponse.
 
Shuffle dimensions of x according to a permutation.
 
Solves systems of linear equations with upper or lower triangular matrices by backsubstitution.
 
Optional attributes for TriangularSolve
Calculate product with tridiagonal matrix.
 
Solves tridiagonal systems of equations.
 
Optional attributes for TridiagonalSolve
Metric that calculates the number of true negatives.
Metric that calculates the number of true positives.
Returns x / y element-wise, rounded towards zero.
 
Initializer that generates a truncated normal distribution.
Outputs random values from a truncated normal distribution.
 
Optional attributes for TruncatedNormal
Returns element-wise remainder of division.
 
String type.
Common interface for all typed tensors.
16-bit unsigned integer tensor type.
8-bit unsigned integer tensor type.
 
Reverses the operation of Batch for a single output Tensor.
 
Optional attributes for Unbatch
A dataset that splits the elements of its input into multiple elements.
A dataset that splits the elements of its input into multiple elements.
 
 
Optional attributes for UnbatchDataset
Gradient of Unbatch.
 
Optional attributes for UnbatchGrad
Uncompresses a compressed dataset element.
 
Decodes each string in input into a sequence of Unicode code points.
 
Optional attributes for UnicodeDecode
Decodes each string in input into a sequence of Unicode code points.
 
Optional attributes for UnicodeDecodeWithOffsets
Encode a tensor of ints into unicode strings.
 
Optional attributes for UnicodeEncode
Determine the script codes of a given tensor of Unicode integer code points.
 
Transcode the input text from a source encoding to a destination encoding.
 
Optional attributes for UnicodeTranscode
Generates labels for candidate sampling with a uniform distribution.
 
Optional attributes for UniformCandidateSampler
Perform dequantization on the quantized Tensor input.
 
Optional attributes for UniformDequantize
Perform quantization on Tensor input.
 
Optional attributes for UniformQuantize
Perform quantized add of quantized Tensor lhs and quantized Tensor rhs to make quantized output.
 
Optional attributes for UniformQuantizedAdd
Perform clip by value on the quantized Tensor operand.
 
Optional attributes for UniformQuantizedClipByValue
Perform quantized convolution of quantized Tensor lhs and quantized Tensor rhs. to make quantized output.
 
Optional attributes for UniformQuantizedConvolution
Perform hybrid quantized convolution of float Tensor lhs and quantized Tensor rhs.
 
Optional attributes for UniformQuantizedConvolutionHybrid
Perform quantized dot of quantized Tensor lhs and quantized Tensor rhs to make quantized output.
 
Optional attributes for UniformQuantizedDot
Perform hybrid quantized dot of float Tensor lhs and quantized Tensor rhs.
 
Optional attributes for UniformQuantizedDotHybrid
 
Describes the dimension numbers for Convolution op.
Describes the dimension numbers for Convolution op.
 
Given quantized tensor input, requantize it with new quantization parameters.
 
Optional attributes for UniformRequantize
Finds unique elements along an axis of a tensor.
 
Creates a dataset that contains the unique elements of input_dataset.
Creates a dataset that contains the unique elements of input_dataset.
 
 
Optional attributes for UniqueDataset
Finds unique elements along an axis of a tensor.
 
Constrains the weights to have unit norm.
Converts an array of flat indices into a tuple of coordinate arrays.
 
The UnsortedSegmentJoin operation
 
Optional attributes for UnsortedSegmentJoin
Computes the maximum along segments of a tensor.
 
Computes the minimum along segments of a tensor.
 
Computes the product along segments of a tensor.
 
Computes the sum along segments of a tensor.
 
Unpacks a given dimension of a rank-R tensor into num rank-(R-1) tensors.
 
Optional attributes for Unstack
Op is similar to a lightweight Dequeue.
 
Optional attributes for Unstage
The UnwrapDatasetVariant operation
 
An op to update the task ID and global core array.
 
Converts all lowercase characters into their respective uppercase replacements.
 
Optional attributes for Upper
Applies upper_bound(sorted_search_values, values) along each row.
 
Protocol buffer representing the values in ControlFlowContext.
Protocol buffer representing the values in ControlFlowContext.
 
Creates a handle to a Variable resource.
 
Optional attributes for VarHandleOp
Holds state in the form of a tensor that persists across steps.
 
Optional attributes for Variable
Indicates how a distributed variable will be aggregated.
Protocol buffer representing a Variable.
Protocol buffer representing a Variable.
 
 
Returns the shape of the variable pointed to by resource.
 
Indicates when a distributed variable will be synced.
Initializer capable of adapting its scale to the shape of weights tensors.
The random distribution to use when initializing the values.
The mode to use for calculating the fan values.
Protocol buffer representing the serialization format of DT_VARIANT tensors.
Protocol buffer representing the serialization format of DT_VARIANT tensors.
 
Checks whether a resource handle-based variable has been initialized.
 
Protobuf type tensorflow.VarLenFeatureProto
Protobuf type tensorflow.VarLenFeatureProto
 
The config for graph verifiers.
The config for graph verifiers.
Protobuf enum tensorflow.VerifierConfig.Toggle
 
 
Version information for a piece of serialized data There are different types of versions for each type of data (GraphDef, etc.), but they all have the same common shape described here.
Version information for a piece of serialized data There are different types of versions for each type of data (GraphDef, etc.), but they all have the same common shape described here.
 
 
Protobuf type tensorflow.WatchdogConfig
Protobuf type tensorflow.WatchdogConfig
 
A minimalist pointer scope only keeping weak references to its elements.
The WeightedFlatMapDataset operation
 
Optional attributes for WeightedFlatMapDataset
Returns locations of nonzero / true values in a tensor.
 
output = input; While (Cond(output)) { output = Body(output) }
Optional attributes for While
Protocol buffer representing a WhileContext object.
Protocol buffer representing a WhileContext object.
 
A Reader that outputs the entire contents of a file as a value.
 
Optional attributes for WholeFileReader
Combines (nests of) input elements into a dataset of (nests of) windows.
 
Optional attributes for WindowDataset
The WindowOp operation
 
Current health status of a worker.
Worker heartbeat op.
 
Protobuf type tensorflow.WorkerHeartbeatRequest
Protobuf type tensorflow.WorkerHeartbeatRequest
 
Protobuf type tensorflow.WorkerHeartbeatResponse
Protobuf type tensorflow.WorkerHeartbeatResponse
 
Indicates the behavior of the worker when an internal error or shutdown signal is received.
The WrapDatasetVariant operation
 
Writes an audio summary.
 
Optional attributes for WriteAudioSummary
Writes contents to the file at input filename.
 
Writes a graph summary.
 
Writes a histogram summary.
 
Writes an image summary.
 
Optional attributes for WriteImageSummary
Writes a serialized proto summary.
 
Writes a scalar summary.
 
Writes a tensor summary.
 
Returns 0 if x == 0, and x / y otherwise, elementwise.
 
A pseudo-op to represent host-side computation in an XLA program.
 
Optional attributes for XlaHostCompute
An API for building xla operations as Ops
An op to receive a tensor from the host. output: the tensor that will be received from the host.
 
An op that receives embedding activations on the TPU.
 
Receives deduplication data (indices and weights) from the embedding core.
 
An op to send a tensor to the host. input: the tensor that will be sent to the host.
 
An op that performs gradient updates of embedding tables.
 
Optional attributes for XlaSendTPUEmbeddingGradients
The XlaSparseCoreAdagrad operation
 
The XlaSparseCoreAdagradMomentum operation
 
The XlaSparseCoreAdam operation
 
The XlaSparseCoreFtrl operation
 
The XlaSparseCoreSgd operation
 
The XlaSparseDenseMatmul operation
 
The XlaSparseDenseMatmulGradWithAdagradAndCsrInput operation
 
The XlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize operation
 
The XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput operation
 
The XlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize operation
 
The XlaSparseDenseMatmulGradWithAdamAndCsrInput operation
 
The XlaSparseDenseMatmulGradWithAdamAndStaticBufferSize operation
 
The XlaSparseDenseMatmulGradWithCsrInput operation
 
The XlaSparseDenseMatmulGradWithFtrlAndCsrInput operation
 
The XlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize operation
 
The XlaSparseDenseMatmulGradWithSgdAndCsrInput operation
 
The XlaSparseDenseMatmulGradWithSgdAndStaticBufferSize operation
 
The XlaSparseDenseMatmulWithCsrInput operation
 
The XlaSparseDenseMatmulWithStaticBufferSize operation
 
Returns 0 if x == 0, and x * log1p(y) otherwise, elementwise.
 
Returns 0 if x == 0, and x * log(y) otherwise, elementwise.
 
 
 
An XEvent is a trace event, optionally annotated with XStats.
An XEvent is a trace event, optionally annotated with XStats.
 
Metadata for an XEvent, corresponds to an event type and is shared by all XEvents with the same metadata_id.
Metadata for an XEvent, corresponds to an event type and is shared by all XEvents with the same metadata_id.
 
 
An XLine is a timeline of trace events (XEvents).
An XLine is a timeline of trace events (XEvents).
 
An XPlane is a container of parallel timelines (XLines), generated by a profiling source or by post-processing one or more XPlanes.
An XPlane is a container of parallel timelines (XLines), generated by a profiling source or by post-processing one or more XPlanes.
 
A container of parallel XPlanes, generated by one or more profiling sources.
A container of parallel XPlanes, generated by one or more profiling sources.
 
An XStat is a named value associated with an XEvent, e.g., a performance counter value, a metric computed by a formula applied over nested XEvents and XStats.
An XStat is a named value associated with an XEvent, e.g., a performance counter value, a metric computed by a formula applied over nested XEvents and XStats.
 
Metadata for an XStat, corresponds to a stat type and is shared by all XStats with the same metadata_id.
Metadata for an XStat, corresponds to a stat type and is shared by all XStats with the same metadata_id.
 
 
Creates an Initializer that sets all values to zero.
An operator creating a constant initialized with zeros of the shape given by `dims`.
Returns a tensor of zeros with the same shape and type as x.
 
Compute the Hurwitz zeta function \(\zeta(x, q)\).
 
Creates a dataset that zips together input_datasets.
 
Optional attributes for ZipDataset