Download and parse MNIST dataset.

Returns an iterator pointing to the given ShapeIndex. REQUIRES: index must exist in the ShapeTree.

helper functions for constructing new subgraph

Add a wrapped input argument to the hint. Args: *args: The input tensor. **kwargs: "name" label "tag" a tag to group

Repeats this dataset `count` times. NOTE: If this dataset is a function of global state (e.g. a random number generator), then different repe

TODO(b/80418076): Move to legacy ops file, update invocations.

Combines consecutive elements of this dataset into batches. The components of the resulting element will have an additional outer dimension,

In a session, computes and returns the value of this variable. This is not a graph construction method, it does not add ops to the graph. Th

Returns true if it is a scalar

Test scope that runs tests on a Tensorflow/XLA device. Uses a compilation_scope() to mark operators to compile. Yields: A scope to app

Returns a symbolic handle to one of the tensors produced by this operation. <p>Warning: Does not check that the type of the tensor matches T. It is r

Creates a generator to extract data from the queue. Skip the data if it is `None`. Yields: The next element in the queue, i.e. a tup

Asserts that message is same as parsed expected_message_ascii. Creates another prototype of message, reads the ascii message into it and then

Return the values in the TensorArray as a stacked `Tensor`. All of the values must have been written and their shapes must all match. If inpu

Creates a `Dataset` by concatenating the given dataset with this dataset. ```python a = Dataset.range(1, 4) # ==> [ 1, 2, 3 ] b = Datase

Returns the <a href="https://www.tensorflow.org/resources/dims_types.html#shape">shape</a> of the Tensor, i.e., the sizes of each dimension. @return

Creates a `Dataset` whose elements are slices of the given tensors. Note that if `tensors` contains a NumPy array, and eager execution is not

Lists registered items. Returns: A list of names of registered objects.

Returns a Tiled sharding attribute. This causes an op to be partially computed on multiple cores in the XLA device. Args: tile_ass

Custom implementation of session() for XLA tests. We override the standard Tensorflow session() since it is too specific to CPU and GPU tests

Generates output predictions for the input samples. Computation is done in batches. Arguments: x: Input samples. It could be:

Slice a RichTextLines object. The object itself is not changed. A sliced instance is returned. Args: begin: (int) Beginning line index

Returns the size of the dimension at the given index.

Performs a mean reduction on `x` across pfor iterations. Note that this currently may not work inside a control flow construct. Args: x

Add a new trackable value.

Returns the `TensorShape` that represents the shape of this tensor. The shape is computed using shape inference functions that are registered

r'''Update params in context.