tensorflow::ops::TemporaryVariable

#include <state_ops.h>

Returns a tensor that may be mutated, but only persists within a single step.

Summary

This is an experimental op for internal use only and it is possible to use this op in unsafe ways. DO NOT USE unless you fully understand the risks.

It is the caller's responsibility to ensure that 'ref' is eventually passed to a matching 'DestroyTemporaryVariable' op after all other uses have completed.

Outputs a ref to the tensor state so it may be read or modified.

E.g. var = state_ops._temporary_variable([1, 2], types.float_) var_name = var.op.name var = state_ops.assign(var, [[4.0, 5.0]]) var = state_ops.assign_add(var, [[6.0, 7.0]]) final = state_ops._destroy_temporary_variable(var, var_name=var_name)

Arguments:

• scope: A Scope object
• shape: The shape of the variable tensor.
• dtype: The type of elements in the variable tensor.

Optional attributes (see Attrs):

• var_name: Overrides the name used for the temporary variable resource. Default value is the name of the 'TemporaryVariable' op (which is guaranteed unique).

Returns:

• Output: A reference to the variable tensor.

Public attributes

::tensorflow::Output ref

Public functions

 TemporaryVariable(
  const ::tensorflow::Scope & scope,
  PartialTensorShape shape,
  DataType dtype
)

 TemporaryVariable(
  const ::tensorflow::Scope & scope,
  PartialTensorShape shape,
  DataType dtype,
  const TemporaryVariable::Attrs & attrs
)

::tensorflow::Node * node() const 

operator::tensorflow::Input() const 

operator::tensorflow::Output() const 

Public static functions

Attrs VarName(
  StringPiece x
)