"""Python wrappers around TensorFlow ops.

This file is MACHINE GENERATED! Do not edit.
"""

import collections

from tensorflow.python import pywrap_tfe as pywrap_tfe
from tensorflow.python.eager import context as _context
from tensorflow.python.eager import core as _core
from tensorflow.python.eager import execute as _execute
from tensorflow.python.framework import dtypes as _dtypes

from tensorflow.python.framework import op_def_registry as _op_def_registry
from tensorflow.python.framework import ops as _ops
from tensorflow.python.framework import op_def_library as _op_def_library
from tensorflow.python.util.deprecation import deprecated_endpoints
from tensorflow.python.util import dispatch as _dispatch
from tensorflow.python.util.tf_export import tf_export

from typing import TypeVar

def assign(ref, value, validate_shape=True, use_locking=True, name=None):
  r"""Update 'ref' by assigning 'value' to it.

  This operation outputs "ref" after the assignment is done.

  This makes it easier to chain operations that need to use the reset value.

  Args:
    ref: A mutable `Tensor`.
      Should be from a `Variable` node. May be uninitialized.
    value: A `Tensor`. Must have the same type as `ref`.
      The value to be assigned to the variable.
    validate_shape: An optional `bool`. Defaults to `True`.
      If true, the operation will validate that the shape

      of 'value' matches the shape of the Tensor being assigned to.  If false,

      'ref' will take on the shape of 'value'.
    use_locking: An optional `bool`. Defaults to `True`.
      If True, the assignment will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("assign op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if validate_shape is None:
    validate_shape = True
  validate_shape = _execute.make_bool(validate_shape, "validate_shape")
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "Assign", ref=ref, value=value, validate_shape=validate_shape,
                  use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "validate_shape",
              _op._get_attr_bool("validate_shape"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "Assign", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

Assign = tf_export("raw_ops.Assign")(_ops.to_raw_op(assign))


def assign_eager_fallback(ref, value, validate_shape, use_locking, name, ctx):
  raise RuntimeError("assign op does not support eager execution. Arg 'output_ref' is a ref.")

def assign_add(ref, value, use_locking=False, name=None):
  r"""Update 'ref' by adding 'value' to it.

  This operation outputs "ref" after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      Should be from a `Variable` node.
    value: A `Tensor`. Must have the same type as `ref`.
      The value to be added to the variable.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the addition will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("assign_add op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "AssignAdd", ref=ref, value=value, use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "AssignAdd", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

AssignAdd = tf_export("raw_ops.AssignAdd")(_ops.to_raw_op(assign_add))


def assign_add_eager_fallback(ref, value, use_locking, name, ctx):
  raise RuntimeError("assign_add op does not support eager execution. Arg 'output_ref' is a ref.")

def assign_sub(ref, value, use_locking=False, name=None):
  r"""Update 'ref' by subtracting 'value' from it.

  This operation outputs "ref" after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      Should be from a `Variable` node.
    value: A `Tensor`. Must have the same type as `ref`.
      The value to be subtracted to the variable.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the subtraction will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("assign_sub op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "AssignSub", ref=ref, value=value, use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "AssignSub", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

AssignSub = tf_export("raw_ops.AssignSub")(_ops.to_raw_op(assign_sub))


def assign_sub_eager_fallback(ref, value, use_locking, name, ctx):
  raise RuntimeError("assign_sub op does not support eager execution. Arg 'output_ref' is a ref.")

def count_up_to(ref, limit, name=None):
  r"""Increments 'ref' until it reaches 'limit'.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `int32`, `int64`.
      Should be from a scalar `Variable` node.
    limit: An `int`.
      If incrementing ref would bring it above limit, instead generates an

      'OutOfRange' error.
    name: A name for the operation (optional).

  Returns:
    A `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("count_up_to op does not support eager execution. Arg 'ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  limit = _execute.make_int(limit, "limit")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "CountUpTo", ref=ref, limit=limit, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("limit", _op._get_attr_int("limit"), "T",
              _op._get_attr_type("T"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "CountUpTo", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

CountUpTo = tf_export("raw_ops.CountUpTo")(_ops.to_raw_op(count_up_to))


def count_up_to_eager_fallback(ref, limit, name, ctx):
  raise RuntimeError("count_up_to op does not support eager execution. Arg 'ref' is a ref.")

def destroy_temporary_variable(ref, var_name, name=None):
  r"""Destroys the temporary variable and returns its final value.

  Sets output to the value of the Tensor pointed to by 'ref', then destroys

  the temporary variable called 'var_name'.

  All other uses of 'ref' *must* have executed before this op.

  This is typically achieved by chaining the ref through each assign op, or by

  using control dependencies.

  

  Outputs the final value of the tensor pointed to by 'ref'.

  Args:
    ref: A mutable `Tensor`. A reference to the temporary variable tensor.
    var_name: A `string`.
      Name of the temporary variable, usually the name of the matching

      'TemporaryVariable' op.
    name: A name for the operation (optional).

  Returns:
    A `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("destroy_temporary_variable op does not support eager execution. Arg 'ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  var_name = _execute.make_str(var_name, "var_name")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "DestroyTemporaryVariable", ref=ref, var_name=var_name, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "var_name",
              _op.get_attr("var_name"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "DestroyTemporaryVariable", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

DestroyTemporaryVariable = tf_export("raw_ops.DestroyTemporaryVariable")(_ops.to_raw_op(destroy_temporary_variable))


def destroy_temporary_variable_eager_fallback(ref, var_name, name, ctx):
  raise RuntimeError("destroy_temporary_variable op does not support eager execution. Arg 'ref' is a ref.")

def is_variable_initialized(ref, name=None):
  r"""Checks whether a tensor has been initialized.

  Outputs boolean scalar indicating whether the tensor has been initialized.

  Args:
    ref: A mutable `Tensor`.
      Should be from a `Variable` node. May be uninitialized.
    name: A name for the operation (optional).

  Returns:
    A `Tensor` of type `bool`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("is_variable_initialized op does not support eager execution. Arg 'ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "IsVariableInitialized", ref=ref, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("dtype", _op._get_attr_type("dtype"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "IsVariableInitialized", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

IsVariableInitialized = tf_export("raw_ops.IsVariableInitialized")(_ops.to_raw_op(is_variable_initialized))


def is_variable_initialized_eager_fallback(ref, name, ctx):
  raise RuntimeError("is_variable_initialized op does not support eager execution. Arg 'ref' is a ref.")

def resource_count_up_to(resource, limit, T, name=None):
  r"""Increments variable pointed to by 'resource' until it reaches 'limit'.

  Args:
    resource: A `Tensor` of type `resource`.
      Should be from a scalar `Variable` node.
    limit: An `int`.
      If incrementing ref would bring it above limit, instead generates an

      'OutOfRange' error.
    T: A `tf.DType` from: `tf.int32, tf.int64`.
    name: A name for the operation (optional).

  Returns:
    A `Tensor` of type `T`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    try:
      _result = pywrap_tfe.TFE_Py_FastPathExecute(
        _ctx, "ResourceCountUpTo", name, resource, "limit", limit, "T", T)
      return _result
    except _core._NotOkStatusException as e:
      _ops.raise_from_not_ok_status(e, name)
    except _core._FallbackException:
      pass
    try:
      return resource_count_up_to_eager_fallback(
          resource, limit=limit, T=T, name=name, ctx=_ctx)
    except _core._SymbolicException:
      pass  # Add nodes to the TensorFlow graph.
  # Add nodes to the TensorFlow graph.
  limit = _execute.make_int(limit, "limit")
  T = _execute.make_type(T, "T")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ResourceCountUpTo", resource=resource, limit=limit, T=T, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("limit", _op._get_attr_int("limit"), "T",
              _op._get_attr_type("T"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ResourceCountUpTo", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ResourceCountUpTo = tf_export("raw_ops.ResourceCountUpTo")(_ops.to_raw_op(resource_count_up_to))


def resource_count_up_to_eager_fallback(resource, limit, T, name, ctx):
  limit = _execute.make_int(limit, "limit")
  T = _execute.make_type(T, "T")
  resource = _ops.convert_to_tensor(resource, _dtypes.resource)
  _inputs_flat = [resource]
  _attrs = ("limit", limit, "T", T)
  _result = _execute.execute(b"ResourceCountUpTo", 1, inputs=_inputs_flat,
                             attrs=_attrs, ctx=ctx, name=name)
  if _execute.must_record_gradient():
    _execute.record_gradient(
        "ResourceCountUpTo", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result


def resource_scatter_nd_add(ref, indices, updates, use_locking=True, name=None):
  r"""Applies sparse addition to individual values or slices in a Variable.

  `ref` is a `Tensor` with rank `P` and `indices` is a `Tensor` of rank `Q`.

  

  `indices` must be integer tensor, containing indices into `ref`.

  It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.

  

  The innermost dimension of `indices` (with length `K`) corresponds to

  indices into elements (if `K = P`) or slices (if `K < P`) along the `K`th

  dimension of `ref`.

  

  `updates` is `Tensor` of rank `Q-1+P-K` with shape:

  

  ```

  [d_0, ..., d_{Q-2}, ref.shape[K], ..., ref.shape[P-1]]

  ```

  

  For example, say we want to add 4 scattered elements to a rank-1 tensor to

  8 elements. In Python, that addition would look like this:

  

  ```python

  ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8], use_resource=True)

  indices = tf.constant([[4], [3], [1], [7]])

  updates = tf.constant([9, 10, 11, 12])

  add = tf.scatter_nd_add(ref, indices, updates)

  with tf.Session() as sess:

    print sess.run(add)

  ```

  

  The resulting update to ref would look like this:

  

      [1, 13, 3, 14, 14, 6, 7, 20]

  

  See `tf.scatter_nd` for more details about how to make updates to

  slices.

  Args:
    ref: A `Tensor` of type `resource`.
      A resource handle. Must be from a VarHandleOp.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`.
      A Tensor. Must have the same type as ref. A tensor of

      values to add to ref.
    use_locking: An optional `bool`. Defaults to `True`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    The created Operation.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    try:
      _result = pywrap_tfe.TFE_Py_FastPathExecute(
        _ctx, "ResourceScatterNdAdd", name, ref, indices, updates,
        "use_locking", use_locking)
      return _result
    except _core._NotOkStatusException as e:
      _ops.raise_from_not_ok_status(e, name)
    except _core._FallbackException:
      pass
    try:
      return resource_scatter_nd_add_eager_fallback(
          ref, indices, updates, use_locking=use_locking, name=name, ctx=_ctx)
    except _core._SymbolicException:
      pass  # Add nodes to the TensorFlow graph.
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ResourceScatterNdAdd", ref=ref, indices=indices, updates=updates,
                                use_locking=use_locking, name=name)
  return _op
ResourceScatterNdAdd = tf_export("raw_ops.ResourceScatterNdAdd")(_ops.to_raw_op(resource_scatter_nd_add))


def resource_scatter_nd_add_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _attr_T, (updates,) = _execute.args_to_matching_eager([updates], ctx, [])
  _attr_Tindices, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ])
  ref = _ops.convert_to_tensor(ref, _dtypes.resource)
  _inputs_flat = [ref, indices, updates]
  _attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "use_locking",
  use_locking)
  _result = _execute.execute(b"ResourceScatterNdAdd", 0, inputs=_inputs_flat,
                             attrs=_attrs, ctx=ctx, name=name)
  _result = None
  return _result


def resource_scatter_nd_max(ref, indices, updates, use_locking=True, name=None):
  r"""TODO: add doc.

  Args:
    ref: A `Tensor` of type `resource`.
      A resource handle. Must be from a VarHandleOp.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`.
      A Tensor. Must have the same type as ref. A tensor of

      values whose element wise max is taken with ref
    use_locking: An optional `bool`. Defaults to `True`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    The created Operation.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    try:
      _result = pywrap_tfe.TFE_Py_FastPathExecute(
        _ctx, "ResourceScatterNdMax", name, ref, indices, updates,
        "use_locking", use_locking)
      return _result
    except _core._NotOkStatusException as e:
      _ops.raise_from_not_ok_status(e, name)
    except _core._FallbackException:
      pass
    try:
      return resource_scatter_nd_max_eager_fallback(
          ref, indices, updates, use_locking=use_locking, name=name, ctx=_ctx)
    except _core._SymbolicException:
      pass  # Add nodes to the TensorFlow graph.
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ResourceScatterNdMax", ref=ref, indices=indices, updates=updates,
                                use_locking=use_locking, name=name)
  return _op
ResourceScatterNdMax = tf_export("raw_ops.ResourceScatterNdMax")(_ops.to_raw_op(resource_scatter_nd_max))


def resource_scatter_nd_max_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _attr_T, (updates,) = _execute.args_to_matching_eager([updates], ctx, [])
  _attr_Tindices, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ])
  ref = _ops.convert_to_tensor(ref, _dtypes.resource)
  _inputs_flat = [ref, indices, updates]
  _attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "use_locking",
  use_locking)
  _result = _execute.execute(b"ResourceScatterNdMax", 0, inputs=_inputs_flat,
                             attrs=_attrs, ctx=ctx, name=name)
  _result = None
  return _result


def resource_scatter_nd_min(ref, indices, updates, use_locking=True, name=None):
  r"""TODO: add doc.

  Args:
    ref: A `Tensor` of type `resource`.
      A resource handle. Must be from a VarHandleOp.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`.
      A Tensor. Must have the same type as ref. A tensor of

      values whose element wise min is taken with ref.
    use_locking: An optional `bool`. Defaults to `True`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    The created Operation.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    try:
      _result = pywrap_tfe.TFE_Py_FastPathExecute(
        _ctx, "ResourceScatterNdMin", name, ref, indices, updates,
        "use_locking", use_locking)
      return _result
    except _core._NotOkStatusException as e:
      _ops.raise_from_not_ok_status(e, name)
    except _core._FallbackException:
      pass
    try:
      return resource_scatter_nd_min_eager_fallback(
          ref, indices, updates, use_locking=use_locking, name=name, ctx=_ctx)
    except _core._SymbolicException:
      pass  # Add nodes to the TensorFlow graph.
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ResourceScatterNdMin", ref=ref, indices=indices, updates=updates,
                                use_locking=use_locking, name=name)
  return _op
ResourceScatterNdMin = tf_export("raw_ops.ResourceScatterNdMin")(_ops.to_raw_op(resource_scatter_nd_min))


def resource_scatter_nd_min_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _attr_T, (updates,) = _execute.args_to_matching_eager([updates], ctx, [])
  _attr_Tindices, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ])
  ref = _ops.convert_to_tensor(ref, _dtypes.resource)
  _inputs_flat = [ref, indices, updates]
  _attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "use_locking",
  use_locking)
  _result = _execute.execute(b"ResourceScatterNdMin", 0, inputs=_inputs_flat,
                             attrs=_attrs, ctx=ctx, name=name)
  _result = None
  return _result


def resource_scatter_nd_sub(ref, indices, updates, use_locking=True, name=None):
  r"""Applies sparse subtraction to individual values or slices in a Variable.

  `ref` is a `Tensor` with rank `P` and `indices` is a `Tensor` of rank `Q`.

  

  `indices` must be integer tensor, containing indices into `ref`.

  It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.

  

  The innermost dimension of `indices` (with length `K`) corresponds to

  indices into elements (if `K = P`) or slices (if `K < P`) along the `K`th

  dimension of `ref`.

  

  `updates` is `Tensor` of rank `Q-1+P-K` with shape:

  

  ```

  [d_0, ..., d_{Q-2}, ref.shape[K], ..., ref.shape[P-1]]

  ```

  

  For example, say we want to subtract 4 scattered elements from a rank-1 tensor

  with 8 elements. In Python, that subtraction would look like this:

  

  ```python

  ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8], use_resource=True)

  indices = tf.constant([[4], [3], [1], [7]])

  updates = tf.constant([9, 10, 11, 12])

  sub = tf.scatter_nd_sub(ref, indices, updates)

  with tf.Session() as sess:

    print sess.run(sub)

  ```

  

  The resulting update to ref would look like this:

  

      [1, -9, 3, -6, -4, 6, 7, -4]

  

  See `tf.scatter_nd` for more details about how to make updates to

  slices.

  Args:
    ref: A `Tensor` of type `resource`.
      A resource handle. Must be from a VarHandleOp.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`.
      A Tensor. Must have the same type as ref. A tensor of

      values to add to ref.
    use_locking: An optional `bool`. Defaults to `True`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    The created Operation.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    try:
      _result = pywrap_tfe.TFE_Py_FastPathExecute(
        _ctx, "ResourceScatterNdSub", name, ref, indices, updates,
        "use_locking", use_locking)
      return _result
    except _core._NotOkStatusException as e:
      _ops.raise_from_not_ok_status(e, name)
    except _core._FallbackException:
      pass
    try:
      return resource_scatter_nd_sub_eager_fallback(
          ref, indices, updates, use_locking=use_locking, name=name, ctx=_ctx)
    except _core._SymbolicException:
      pass  # Add nodes to the TensorFlow graph.
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ResourceScatterNdSub", ref=ref, indices=indices, updates=updates,
                                use_locking=use_locking, name=name)
  return _op
ResourceScatterNdSub = tf_export("raw_ops.ResourceScatterNdSub")(_ops.to_raw_op(resource_scatter_nd_sub))


def resource_scatter_nd_sub_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _attr_T, (updates,) = _execute.args_to_matching_eager([updates], ctx, [])
  _attr_Tindices, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ])
  ref = _ops.convert_to_tensor(ref, _dtypes.resource)
  _inputs_flat = [ref, indices, updates]
  _attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "use_locking",
  use_locking)
  _result = _execute.execute(b"ResourceScatterNdSub", 0, inputs=_inputs_flat,
                             attrs=_attrs, ctx=ctx, name=name)
  _result = None
  return _result


def resource_scatter_nd_update(ref, indices, updates, use_locking=True, name=None):
  r"""Applies sparse `updates` to individual values or slices within a given

  variable according to `indices`.

  

  `ref` is a `Tensor` with rank `P` and `indices` is a `Tensor` of rank `Q`.

  

  `indices` must be integer tensor, containing indices into `ref`.

  It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.

  

  The innermost dimension of `indices` (with length `K`) corresponds to

  indices into elements (if `K = P`) or slices (if `K < P`) along the `K`th

  dimension of `ref`.

  

  `updates` is `Tensor` of rank `Q-1+P-K` with shape:

  

  ```

  [d_0, ..., d_{Q-2}, ref.shape[K], ..., ref.shape[P-1]].

  ```

  

  For example, say we want to update 4 scattered elements to a rank-1 tensor to

  8 elements. In Python, that update would look like this:

  

  ```python

      ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8])

      indices = tf.constant([[4], [3], [1] ,[7]])

      updates = tf.constant([9, 10, 11, 12])

      update = tf.scatter_nd_update(ref, indices, updates)

      with tf.Session() as sess:

        print sess.run(update)

  ```

  

  The resulting update to ref would look like this:

  

      [1, 11, 3, 10, 9, 6, 7, 12]

  

  See `tf.scatter_nd` for more details about how to make updates to

  slices.

  Args:
    ref: A `Tensor` of type `resource`.
      A resource handle. Must be from a VarHandleOp.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`.
      A Tensor. Must have the same type as ref. A tensor of updated

      values to add to ref.
    use_locking: An optional `bool`. Defaults to `True`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    The created Operation.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    try:
      _result = pywrap_tfe.TFE_Py_FastPathExecute(
        _ctx, "ResourceScatterNdUpdate", name, ref, indices, updates,
        "use_locking", use_locking)
      return _result
    except _core._NotOkStatusException as e:
      _ops.raise_from_not_ok_status(e, name)
    except _core._FallbackException:
      pass
    try:
      return resource_scatter_nd_update_eager_fallback(
          ref, indices, updates, use_locking=use_locking, name=name, ctx=_ctx)
    except _core._SymbolicException:
      pass  # Add nodes to the TensorFlow graph.
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ResourceScatterNdUpdate", ref=ref, indices=indices, updates=updates,
                                   use_locking=use_locking, name=name)
  return _op
ResourceScatterNdUpdate = tf_export("raw_ops.ResourceScatterNdUpdate")(_ops.to_raw_op(resource_scatter_nd_update))


def resource_scatter_nd_update_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _attr_T, (updates,) = _execute.args_to_matching_eager([updates], ctx, [])
  _attr_Tindices, (indices,) = _execute.args_to_matching_eager([indices], ctx, [_dtypes.int32, _dtypes.int64, ])
  ref = _ops.convert_to_tensor(ref, _dtypes.resource)
  _inputs_flat = [ref, indices, updates]
  _attrs = ("T", _attr_T, "Tindices", _attr_Tindices, "use_locking",
  use_locking)
  _result = _execute.execute(b"ResourceScatterNdUpdate", 0,
                             inputs=_inputs_flat, attrs=_attrs, ctx=ctx,
                             name=name)
  _result = None
  return _result


def scatter_add(ref, indices, updates, use_locking=False, name=None):
  r"""Adds sparse updates to a variable reference.

  This operation computes

  

      # Scalar indices

      ref[indices, ...] += updates[...]

  

      # Vector indices (for each i)

      ref[indices[i], ...] += updates[i, ...]

  

      # High rank indices (for each i, ..., j)

      ref[indices[i, ..., j], ...] += updates[i, ..., j, ...]

  

  This operation outputs `ref` after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  

  Duplicate entries are handled correctly: if multiple `indices` reference

  the same location, their contributions add.

  

  Requires `updates.shape = indices.shape + ref.shape[1:]` or `updates.shape = []`.

  

  <div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">

  <img style="width:100%" src="https://www.tensorflow.org/images/ScatterAdd.png" alt>

  </div>

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      Should be from a `Variable` node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A tensor of indices into the first dimension of `ref`.
    updates: A `Tensor`. Must have the same type as `ref`.
      A tensor of updated values to add to `ref`.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the addition will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_add op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterAdd", ref=ref, indices=indices, updates=updates,
                      use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterAdd", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterAdd = tf_export("raw_ops.ScatterAdd")(_ops.to_raw_op(scatter_add))


def scatter_add_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_add op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_div(ref, indices, updates, use_locking=False, name=None):
  r"""Divides a variable reference by sparse updates.

  This operation computes

  

  ```python

      # Scalar indices

      ref[indices, ...] /= updates[...]

  

      # Vector indices (for each i)

      ref[indices[i], ...] /= updates[i, ...]

  

      # High rank indices (for each i, ..., j)

      ref[indices[i, ..., j], ...] /= updates[i, ..., j, ...]

  ```

  

  This operation outputs `ref` after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  

  Duplicate entries are handled correctly: if multiple `indices` reference

  the same location, their contributions divide.

  

  Requires `updates.shape = indices.shape + ref.shape[1:]` or `updates.shape = []`.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      Should be from a `Variable` node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A tensor of indices into the first dimension of `ref`.
    updates: A `Tensor`. Must have the same type as `ref`.
      A tensor of values that `ref` is divided by.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the operation will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_div op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterDiv", ref=ref, indices=indices, updates=updates,
                      use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterDiv", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterDiv = tf_export("raw_ops.ScatterDiv")(_ops.to_raw_op(scatter_div))


def scatter_div_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_div op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_max(ref, indices, updates, use_locking=False, name=None):
  r"""Reduces sparse updates into a variable reference using the `max` operation.

  This operation computes

  

      # Scalar indices

      ref[indices, ...] = max(ref[indices, ...], updates[...])

  

      # Vector indices (for each i)

      ref[indices[i], ...] = max(ref[indices[i], ...], updates[i, ...])

  

      # High rank indices (for each i, ..., j)

      ref[indices[i, ..., j], ...] = max(ref[indices[i, ..., j], ...], updates[i, ..., j, ...])

  

  This operation outputs `ref` after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  

  Duplicate entries are handled correctly: if multiple `indices` reference

  the same location, their contributions combine.

  

  Requires `updates.shape = indices.shape + ref.shape[1:]` or `updates.shape = []`.

  

  <div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">

  <img style="width:100%" src="https://www.tensorflow.org/images/ScatterAdd.png" alt>

  </div>

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `half`, `bfloat16`, `float32`, `float64`, `int32`, `int64`.
      Should be from a `Variable` node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A tensor of indices into the first dimension of `ref`.
    updates: A `Tensor`. Must have the same type as `ref`.
      A tensor of updated values to reduce into `ref`.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the update will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_max op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterMax", ref=ref, indices=indices, updates=updates,
                      use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterMax", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterMax = tf_export("raw_ops.ScatterMax")(_ops.to_raw_op(scatter_max))


def scatter_max_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_max op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_min(ref, indices, updates, use_locking=False, name=None):
  r"""Reduces sparse updates into a variable reference using the `min` operation.

  This operation computes

  

      # Scalar indices

      ref[indices, ...] = min(ref[indices, ...], updates[...])

  

      # Vector indices (for each i)

      ref[indices[i], ...] = min(ref[indices[i], ...], updates[i, ...])

  

      # High rank indices (for each i, ..., j)

      ref[indices[i, ..., j], ...] = min(ref[indices[i, ..., j], ...], updates[i, ..., j, ...])

  

  This operation outputs `ref` after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  

  Duplicate entries are handled correctly: if multiple `indices` reference

  the same location, their contributions combine.

  

  Requires `updates.shape = indices.shape + ref.shape[1:]` or `updates.shape = []`.

  

  <div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">

  <img style="width:100%" src="https://www.tensorflow.org/images/ScatterAdd.png" alt>

  </div>

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `half`, `bfloat16`, `float32`, `float64`, `int32`, `int64`.
      Should be from a `Variable` node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A tensor of indices into the first dimension of `ref`.
    updates: A `Tensor`. Must have the same type as `ref`.
      A tensor of updated values to reduce into `ref`.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the update will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_min op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterMin", ref=ref, indices=indices, updates=updates,
                      use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterMin", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterMin = tf_export("raw_ops.ScatterMin")(_ops.to_raw_op(scatter_min))


def scatter_min_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_min op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_mul(ref, indices, updates, use_locking=False, name=None):
  r"""Multiplies sparse updates into a variable reference.

  This operation computes

  

  ```python

      # Scalar indices

      ref[indices, ...] *= updates[...]

  

      # Vector indices (for each i)

      ref[indices[i], ...] *= updates[i, ...]

  

      # High rank indices (for each i, ..., j)

      ref[indices[i, ..., j], ...] *= updates[i, ..., j, ...]

  ```

  

  This operation outputs `ref` after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  

  Duplicate entries are handled correctly: if multiple `indices` reference

  the same location, their contributions multiply.

  

  Requires `updates.shape = indices.shape + ref.shape[1:]` or `updates.shape = []`.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      Should be from a `Variable` node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A tensor of indices into the first dimension of `ref`.
    updates: A `Tensor`. Must have the same type as `ref`.
      A tensor of updated values to multiply to `ref`.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the operation will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_mul op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterMul", ref=ref, indices=indices, updates=updates,
                      use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterMul", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterMul = tf_export("raw_ops.ScatterMul")(_ops.to_raw_op(scatter_mul))


def scatter_mul_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_mul op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_nd_add(ref, indices, updates, use_locking=False, name=None):
  r"""Applies sparse addition to individual values or slices in a Variable.

  `ref` is a `Tensor` with rank `P` and `indices` is a `Tensor` of rank `Q`.

  

  `indices` must be integer tensor, containing indices into `ref`.

  It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.

  

  The innermost dimension of `indices` (with length `K`) corresponds to

  indices into elements (if `K = P`) or slices (if `K < P`) along the `K`th

  dimension of `ref`.

  

  `updates` is `Tensor` of rank `Q-1+P-K` with shape:

  

  ```

  [d_0, ..., d_{Q-2}, ref.shape[K], ..., ref.shape[P-1]]

  ```

  

  For example, say we want to add 4 scattered elements to a rank-1 tensor to

  8 elements. In Python, that addition would look like this:

  

  ```python

  ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8])

  indices = tf.constant([[4], [3], [1], [7]])

  updates = tf.constant([9, 10, 11, 12])

  add = tf.scatter_nd_add(ref, indices, updates)

  with tf.Session() as sess:

    print sess.run(add)

  ```

  

  The resulting update to ref would look like this:

  

      [1, 13, 3, 14, 14, 6, 7, 20]

  

  See `tf.scatter_nd` for more details about how to make updates to

  slices.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      A mutable Tensor. Should be from a Variable node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`. Must have the same type as `ref`.
      A Tensor. Must have the same type as ref. A tensor of updated values

      to add to ref.
    use_locking: An optional `bool`. Defaults to `False`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_nd_add op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterNdAdd", ref=ref, indices=indices, updates=updates,
                        use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterNdAdd", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterNdAdd = tf_export("raw_ops.ScatterNdAdd")(_ops.to_raw_op(scatter_nd_add))


def scatter_nd_add_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_nd_add op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_nd_max(ref, indices, updates, use_locking=False, name=None):
  r"""Computes element-wise maximum.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      A mutable Tensor. Should be from a Variable node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`. Must have the same type as `ref`.
      A Tensor. Must have the same type as ref. A tensor of updated values

      to add to ref.
    use_locking: An optional `bool`. Defaults to `False`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_nd_max op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterNdMax", ref=ref, indices=indices, updates=updates,
                        use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterNdMax", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterNdMax = tf_export("raw_ops.ScatterNdMax")(_ops.to_raw_op(scatter_nd_max))


def scatter_nd_max_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_nd_max op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_nd_min(ref, indices, updates, use_locking=False, name=None):
  r"""Computes element-wise minimum.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      A mutable Tensor. Should be from a Variable node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`. Must have the same type as `ref`.
      A Tensor. Must have the same type as ref. A tensor of updated values

      to add to ref.
    use_locking: An optional `bool`. Defaults to `False`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_nd_min op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterNdMin", ref=ref, indices=indices, updates=updates,
                        use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterNdMin", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterNdMin = tf_export("raw_ops.ScatterNdMin")(_ops.to_raw_op(scatter_nd_min))


def scatter_nd_min_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_nd_min op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_nd_sub(ref, indices, updates, use_locking=False, name=None):
  r"""Applies sparse subtraction to individual values or slices in a Variable.

  within a given variable according to `indices`.

  

  `ref` is a `Tensor` with rank `P` and `indices` is a `Tensor` of rank `Q`.

  

  `indices` must be integer tensor, containing indices into `ref`.

  It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.

  

  The innermost dimension of `indices` (with length `K`) corresponds to

  indices into elements (if `K = P`) or slices (if `K < P`) along the `K`th

  dimension of `ref`.

  

  `updates` is `Tensor` of rank `Q-1+P-K` with shape:

  

  ```

  [d_0, ..., d_{Q-2}, ref.shape[K], ..., ref.shape[P-1]]

  ```

  

  For example, say we want to subtract 4 scattered elements from a rank-1 tensor

  with 8 elements. In Python, that subtraction would look like this:

  

  ```python

  ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8])

  indices = tf.constant([[4], [3], [1], [7]])

  updates = tf.constant([9, 10, 11, 12])

  sub = tf.scatter_nd_sub(ref, indices, updates)

  with tf.Session() as sess:

    print sess.run(sub)

  ```

  

  The resulting update to ref would look like this:

  

      [1, -9, 3, -6, -4, 6, 7, -4]

  

  See `tf.scatter_nd` for more details about how to make updates to

  slices.

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      A mutable Tensor. Should be from a Variable node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`. Must have the same type as `ref`.
      A Tensor. Must have the same type as ref. A tensor of updated values

      to subtract from ref.
    use_locking: An optional `bool`. Defaults to `False`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_nd_sub op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterNdSub", ref=ref, indices=indices, updates=updates,
                        use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterNdSub", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterNdSub = tf_export("raw_ops.ScatterNdSub")(_ops.to_raw_op(scatter_nd_sub))


def scatter_nd_sub_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_nd_sub op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_nd_update(ref, indices, updates, use_locking=True, name=None):
  r"""Applies sparse `updates` to individual values or slices within a given

  variable according to `indices`.

  

  `ref` is a `Tensor` with rank `P` and `indices` is a `Tensor` of rank `Q`.

  

  `indices` must be integer tensor, containing indices into `ref`.

  It must be shape \\([d_0, ..., d_{Q-2}, K]\\) where `0 < K <= P`.

  

  The innermost dimension of `indices` (with length `K`) corresponds to

  indices into elements (if `K = P`) or slices (if `K < P`) along the `K`th

  dimension of `ref`.

  

  `updates` is `Tensor` of rank `Q-1+P-K` with shape:

  

  $$[d_0, ..., d_{Q-2}, ref.shape[K], ..., ref.shape[P-1]].$$

  

  For example, say we want to update 4 scattered elements to a rank-1 tensor to

  8 elements. In Python, that update would look like this:

  

  ```python

      ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8])

      indices = tf.constant([[4], [3], [1] ,[7]])

      updates = tf.constant([9, 10, 11, 12])

      update = tf.scatter_nd_update(ref, indices, updates)

      with tf.Session() as sess:

        print sess.run(update)

  ```

  

  The resulting update to ref would look like this:

  

      [1, 11, 3, 10, 9, 6, 7, 12]

  

  See `tf.scatter_nd` for more details about how to make updates to

  slices.

  

  See also `tf.scatter_update` and `tf.batch_scatter_update`.

  Args:
    ref: A mutable `Tensor`.
      A mutable Tensor. Should be from a Variable node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A Tensor. Must be one of the following types: int32, int64.

      A tensor of indices into ref.
    updates: A `Tensor`. Must have the same type as `ref`.
      A Tensor. Must have the same type as ref. A tensor of updated

      values to add to ref.
    use_locking: An optional `bool`. Defaults to `True`.
      An optional bool. Defaults to True. If True, the assignment will

      be protected by a lock; otherwise the behavior is undefined,

      but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_nd_update op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterNdUpdate", ref=ref, indices=indices, updates=updates,
                           use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterNdUpdate", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterNdUpdate = tf_export("raw_ops.ScatterNdUpdate")(_ops.to_raw_op(scatter_nd_update))


def scatter_nd_update_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_nd_update op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_sub(ref, indices, updates, use_locking=False, name=None):
  r"""Subtracts sparse updates to a variable reference.

  ```python

      # Scalar indices

      ref[indices, ...] -= updates[...]

  

      # Vector indices (for each i)

      ref[indices[i], ...] -= updates[i, ...]

  

      # High rank indices (for each i, ..., j)

      ref[indices[i, ..., j], ...] -= updates[i, ..., j, ...]

  ```

  

  This operation outputs `ref` after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  

  Duplicate entries are handled correctly: if multiple `indices` reference

  the same location, their (negated) contributions add.

  

  Requires `updates.shape = indices.shape + ref.shape[1:]` or `updates.shape = []`.

  

  <div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">

  <img style="width:100%" src="https://www.tensorflow.org/images/ScatterSub.png" alt>

  </div>

  Args:
    ref: A mutable `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, `int64`, `qint8`, `quint8`, `qint32`, `bfloat16`, `qint16`, `quint16`, `uint16`, `complex128`, `half`, `uint32`, `uint64`.
      Should be from a `Variable` node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A tensor of indices into the first dimension of `ref`.
    updates: A `Tensor`. Must have the same type as `ref`.
      A tensor of updated values to subtract from `ref`.
    use_locking: An optional `bool`. Defaults to `False`.
      If True, the subtraction will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_sub op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = False
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterSub", ref=ref, indices=indices, updates=updates,
                      use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterSub", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterSub = tf_export("raw_ops.ScatterSub")(_ops.to_raw_op(scatter_sub))


def scatter_sub_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_sub op does not support eager execution. Arg 'output_ref' is a ref.")

def scatter_update(ref, indices, updates, use_locking=True, name=None):
  r"""Applies sparse updates to a variable reference.

  This operation computes

  

  ```python

      # Scalar indices

      ref[indices, ...] = updates[...]

  

      # Vector indices (for each i)

      ref[indices[i], ...] = updates[i, ...]

  

      # High rank indices (for each i, ..., j)

      ref[indices[i, ..., j], ...] = updates[i, ..., j, ...]

  ```

  

  This operation outputs `ref` after the update is done.

  This makes it easier to chain operations that need to use the reset value.

  

  If values in `ref` is to be updated more than once, because there are

  duplicate entries in `indices`, the order at which the updates happen

  for each value is undefined.

  

  Requires `updates.shape = indices.shape + ref.shape[1:]` or `updates.shape = []`.

  

  <div style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;">

  <img style="width:100%" src="https://www.tensorflow.org/images/ScatterUpdate.png" alt>

  </div>

  

  See also `tf.batch_scatter_update` and `tf.scatter_nd_update`.

  Args:
    ref: A mutable `Tensor`. Should be from a `Variable` node.
    indices: A `Tensor`. Must be one of the following types: `int32`, `int64`.
      A tensor of indices into the first dimension of `ref`.
    updates: A `Tensor`. Must have the same type as `ref`.
      A tensor of updated values to store in `ref`.
    use_locking: An optional `bool`. Defaults to `True`.
      If True, the assignment will be protected by a lock;

      otherwise the behavior is undefined, but may exhibit less contention.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor`. Has the same type as `ref`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("scatter_update op does not support eager execution. Arg 'output_ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  if use_locking is None:
    use_locking = True
  use_locking = _execute.make_bool(use_locking, "use_locking")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "ScatterUpdate", ref=ref, indices=indices, updates=updates,
                         use_locking=use_locking, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("T", _op._get_attr_type("T"), "Tindices",
              _op._get_attr_type("Tindices"), "use_locking",
              _op._get_attr_bool("use_locking"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "ScatterUpdate", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

ScatterUpdate = tf_export("raw_ops.ScatterUpdate")(_ops.to_raw_op(scatter_update))


def scatter_update_eager_fallback(ref, indices, updates, use_locking, name, ctx):
  raise RuntimeError("scatter_update op does not support eager execution. Arg 'output_ref' is a ref.")

def temporary_variable(shape, dtype, var_name="", name=None):
  r"""Returns a tensor that may be mutated, but only persists within a single step.

  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)

  Args:
    shape: A `tf.TensorShape` or list of `ints`.
      The shape of the variable tensor.
    dtype: A `tf.DType`. The type of elements in the variable tensor.
    var_name: An optional `string`. Defaults to `""`.
      Overrides the name used for the temporary variable resource. Default

      value is the name of the 'TemporaryVariable' op (which is guaranteed unique).
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor` of type `dtype`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("temporary_variable op does not support eager execution. Arg 'ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  shape = _execute.make_shape(shape, "shape")
  dtype = _execute.make_type(dtype, "dtype")
  if var_name is None:
    var_name = ""
  var_name = _execute.make_str(var_name, "var_name")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "TemporaryVariable", shape=shape, dtype=dtype, var_name=var_name,
                             name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("shape", _op.get_attr("shape"), "dtype",
              _op._get_attr_type("dtype"), "var_name",
              _op.get_attr("var_name"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "TemporaryVariable", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

TemporaryVariable = tf_export("raw_ops.TemporaryVariable")(_ops.to_raw_op(temporary_variable))


def temporary_variable_eager_fallback(shape, dtype, var_name, name, ctx):
  raise RuntimeError("temporary_variable op does not support eager execution. Arg 'ref' is a ref.")

def variable(shape, dtype, container="", shared_name="", name=None):
  r"""Use VariableV2 instead.

  Args:
    shape: A `tf.TensorShape` or list of `ints`.
    dtype: A `tf.DType`.
    container: An optional `string`. Defaults to `""`.
    shared_name: An optional `string`. Defaults to `""`.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor` of type `dtype`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("variable op does not support eager execution. Arg 'ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  shape = _execute.make_shape(shape, "shape")
  dtype = _execute.make_type(dtype, "dtype")
  if container is None:
    container = ""
  container = _execute.make_str(container, "container")
  if shared_name is None:
    shared_name = ""
  shared_name = _execute.make_str(shared_name, "shared_name")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "Variable", shape=shape, dtype=dtype, container=container,
                    shared_name=shared_name, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("shape", _op.get_attr("shape"), "dtype",
              _op._get_attr_type("dtype"), "container",
              _op.get_attr("container"), "shared_name",
              _op.get_attr("shared_name"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "Variable", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

Variable = tf_export("raw_ops.Variable")(_ops.to_raw_op(variable))


def variable_eager_fallback(shape, dtype, container, shared_name, name, ctx):
  raise RuntimeError("variable op does not support eager execution. Arg 'ref' is a ref.")

def variable_v2(shape, dtype, container="", shared_name="", name=None):
  r"""Holds state in the form of a tensor that persists across steps.

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

  TODO(zhifengc/mrry): Adds a pointer to a more detail document

  about sharing states in tensorflow.

  Args:
    shape: A `tf.TensorShape` or list of `ints`.
      The shape of the variable tensor.
    dtype: A `tf.DType`. The type of elements in the variable tensor.
    container: An optional `string`. Defaults to `""`.
      If non-empty, this variable is placed in the given container.

      Otherwise, a default container is used.
    shared_name: An optional `string`. Defaults to `""`.
      If non-empty, this variable is named in the given bucket

      with this shared_name. Otherwise, the node name is used instead.
    name: A name for the operation (optional).

  Returns:
    A mutable `Tensor` of type `dtype`.
  """
  _ctx = _context._context or _context.context()
  tld = _ctx._thread_local_data
  if tld.is_eager:
    raise RuntimeError("variable_v2 op does not support eager execution. Arg 'ref' is a ref.")
  # Add nodes to the TensorFlow graph.
  shape = _execute.make_shape(shape, "shape")
  dtype = _execute.make_type(dtype, "dtype")
  if container is None:
    container = ""
  container = _execute.make_str(container, "container")
  if shared_name is None:
    shared_name = ""
  shared_name = _execute.make_str(shared_name, "shared_name")
  _, _, _op, _outputs = _op_def_library._apply_op_helper(
        "VariableV2", shape=shape, dtype=dtype, container=container,
                      shared_name=shared_name, name=name)
  _result = _outputs[:]
  if _execute.must_record_gradient():
    _attrs = ("shape", _op.get_attr("shape"), "dtype",
              _op._get_attr_type("dtype"), "container",
              _op.get_attr("container"), "shared_name",
              _op.get_attr("shared_name"))
    _inputs_flat = _op.inputs
    _execute.record_gradient(
        "VariableV2", _inputs_flat, _attrs, _result)
  _result, = _result
  return _result

VariableV2 = tf_export("raw_ops.VariableV2")(_ops.to_raw_op(variable_v2))


def variable_v2_eager_fallback(shape, dtype, container, shared_name, name, ctx):
  raise RuntimeError("variable_v2 op does not support eager execution. Arg 'ref' is a ref.")