Element-wise square root. Arguments: x: Tensor or variable. Returns: A tensor.

Return a string representation of this `DeviceSpec`. Returns: a string of the form /job:<name>/replica:<id>/task:<id>/device:<device_

\brief Binds signed 64-bit integer to 1-indexed query parameter.

Convenience wrapper to create Numpy arrays with a np.float64 dtype.

A context manager that captures the writes to a given stream. This context manager captures all writes to a given stream inside of a `Capture

Returns a singleton context object.

Returns true if the given dimension is dynamically-sized.

Computes the Kullback--Leibler divergence. Denote this distribution (`self`) by `p` and the `other` distribution by `q`. Assuming `p, q` are

Returns the number of elements produced by this itertaor.

DataType returns the scalar datatype of the Tensor.

Determines whether tensor is constant.

Name returns the name of the operation.

Constructs <tf_related_part> for the ILayer name as <tf_node_def_name>_<sub_op_name>_<sub_op_instance> and callSetLayerNameHelper to set the name for

Tests whether the CFG contains the specified edges.

Assert the scope contains specific used, modified & created variables.

Compares two `GraphDump` monitors and returns differences. Args: other_dump: Another `GraphDump` monitor. step: `int`, step to compar

Mock version of api.converted_call.

Merge args across replicas and run `merge_fn` in a cross-replica context. This allows communication and coordination when there are multiple call

The `c` in: `Y = g(X) = (1 + X * c)**(1 / c)`.

Returns predictions for given features. By default, returns predicted scores. But this default will be dropped soon. Users should either pass

Reverses output of transform back to text. Args: x: iterator or matrix of integers. Document representation in bytes. Yields: It

Updates the state for this split handler running the scheduled I/O. Args: stamp_token: An int32 scalar tensor containing the current stamp

Waits for ops dispatched in this executor to finish.

Checks that the number of inputs match, and enforces that the inputs marked as weights are constant. Inputs are allowed to be both weight and tensor.

Convert a TensorShape into the equivalent XLA Shape proto.

Gets the number of computations in this module.

Exits the current frame to its parent frame. Exit makes its input `data` available to the parent frame. Args: data: The tensor to be made av

Load the TensorFlowLite runtime C library. @hide

Main Dense-Tensor-In-Dense-Tensor-Out (DTIDTO) map method. Should be implemented by subclasses. Args: input_tensor: The dense input ten

Create a Runner to execute graph operations and evaluate Tensors.