Holds a tensor or tensor reference. For tensor references, we need a mutex to prevent concurrent access to the tensor.

This is a helper for creating a Node and adding it to a Graph. Internally, it uses a NodeDefBuilder to automatically set attrs that can be inferred fr

A description of a single command line flag, holding its name, type, usage text, and a pointer to the corresponding variable.

Represents the output of 'node' at 'index'.

@ingroup core Denotes success or failure of a call in Tensorflow.

A shape describes the number of dimensions in a array, the bounds of each dimension, and the primitive component type. For tuples, shape describes the

Initializer base class: all initializers inherit from this class.

See the [Variables Guide](https://tensorflow.org/guide/variables). A variable maintains state in the graph across calls to `run()`. You add a var

Shape rank and dimensions are accessed through InferenceContext.

An OpTestBuilder is a graph builder class that takes as input an operator to test, its inputs and attributes, and builds a graph that executes the ope

Builder class passed to the REGISTER_OP() macro.

Key formats for accepting/returning values.

A tensorflow Op may need access to different kinds of memory that are not simply a function of the device to which the Op has been assigned. For exam

Handles caching of transformations while building the model. `_FeatureColumn` specifies how to digest an input column to the network. Some featur

A helper class to make AttrSlice from initializer lists

Identifier for a tensor within a step. first == operation_name, second == output_index Note: does not own backing storage for name.

A DeviceType is just a string, but we wrap it up in a class to give some type checking as we're passing these around

A description of a single command line flag, holding its name, type, usage text, and a pointer to the corresponding variable.

Represents a qualified name.

Represents runtime information needed to construct a dataset.

Scanner provides simplified string parsing, in which a string is parsed as a series of scanning calls (e.g. One, Any, Many, OneLiteral, Eos), and then

For the purpose of control flow, every tensor produced by TensorFlow is conceptually tagged by a 'FrameAndIter'. FrameAndIter consists of a 'frame_id'

Represents a TRT-style input to a TF node, it can be either a nvinfer1::ITensor, or TRT_ShapedWeights which is compile-time constant. TODO(laigd): ma

A pass which performs algebraic simplifications.

Specifies the ndim, dtype and shape of every input to a layer. Every layer should expose (if appropriate) an `input_spec` attribute: a list of in

Interpreter interface for TensorFlow Lite Models. This makes the TensorFlow Lite interpreter accessible in Python. It is possible to use this int

Non-assignable values should work.

A ShapeLayout object encapsulates the layout of a particular shape (including tuples). This differs from the Layout proto which describes the layout o

Default time before reporting failure: ~100 seconds.

MetricSpec connects a model to metric functions. THIS CLASS IS DEPRECATED. See [contrib/learn/README.md](https://www.tensorflow.org/code/tensorfl

Mock subclass of CursesUI that bypasses actual terminal manipulations.

SessionOptions contains configuration information for a session.

Represents the type of the elements in a `Tensor`. The following `DType` objects are defined: * `tf.float16`: 16-bit half-precision floating-poi

Attributes for a single allocation call. Different calls to the same allocator could potentially have different allocation attributes.

An index for specifying a particular nested subshape within a shape. Used in ShapeUtil::GetSubshape and other interfaces. Shapes are recursive data st

Subclasses the wrapper class for testing. Overrides its CLI-related methods for headless testing environments. Inserts observer variables for ass

Optimizer that implements the RMSProp algorithm. See the [paper](http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf).

Compute `Y = g(X; shift, scale) = scale @ X + shift`. Here `scale = c * I + diag(D1) + tril(L) + V @ diag(D2) @ V.T`. In TF parlance, the `scale