Python tensorflow.python.ops.array_ops.shape_internal() Examples

def ZerosLikeOutsideLoop(op, index):
  """Create zeros_like for the specified output of an op."""
  val = op.outputs[index]
  if not IsSwitch(op):
    return array_ops.zeros_like(val, optimize=False)
  else:
    op_ctxt = op._get_control_flow_context()
    if op_ctxt:
      # We are in a cond context. Use a switch to create zeros only when needed.
      pred = op_ctxt.pred
      branch = op_ctxt.branch
      switch_val = switch(op.inputs[0], pred)[1 - branch]
      zeros_shape = array_ops.shape_internal(switch_val, optimize=False)
      return array_ops.zeros(zeros_shape, dtype=val.dtype)
    else:
      return array_ops.zeros_like(val, optimize=False) 

def _as_indexed_slices(x, optimize=True):
  """Convert 'x' to IndexedSlices.

  Convert a dense Tensor to a block-sparse IndexedSlices.

  Args:
    x: Either a Tensor object, or an IndexedSlices object.
    optimize: if true, attempt to optimize the conversion of 'x'.

  Returns:
    An IndexedSlices object.

  Raises:
    TypeError: If 'x' is not a Tensor or an IndexedSlices object.
  """
  # TODO(touts): op_scope
  if not isinstance(x, (ops.Tensor, ops.IndexedSlices)):
    raise TypeError("Not a Tensor or IndexedSlices: %s" % type(x))
  if isinstance(x, ops.IndexedSlices):
    return x
  x_shape = array_ops.shape_internal(x, optimize=optimize)
  return ops.IndexedSlices(x, range(0, x_shape[0]), x_shape) 

def ZerosLikeOutsideLoop(op, index):
  """Create zeros_like for the specified output of an op."""
  val = op.outputs[index]
  if not IsSwitch(op):
    return array_ops.zeros_like(val, optimize=False)
  else:
    op_ctxt = op._get_control_flow_context()
    if op_ctxt:
      # We are in a cond context. Use a switch to create zeros only when needed.
      pred = op_ctxt.pred
      branch = op_ctxt.branch
      switch_val = switch(op.inputs[0], pred)[1 - branch]
      zeros_shape = array_ops.shape_internal(switch_val, optimize=False)
      return array_ops.zeros(zeros_shape, dtype=val.dtype)
    else:
      return array_ops.zeros_like(val, optimize=False) 

def _as_indexed_slices(x, optimize=True):
  """Convert 'x' to IndexedSlices.

  Convert a dense Tensor to a block-sparse IndexedSlices.

  Args:
    x: Either a Tensor object, or an IndexedSlices object.
    optimize: if true, attempt to optimize the conversion of 'x'.

  Returns:
    An IndexedSlices object.

  Raises:
    TypeError: If 'x' is not a Tensor or an IndexedSlices object.
  """
  # TODO(touts): op_scope
  if not isinstance(x, (ops.Tensor, ops.IndexedSlices)):
    raise TypeError("Not a Tensor or IndexedSlices: %s" % type(x))
  if isinstance(x, ops.IndexedSlices):
    return x
  x_shape = array_ops.shape_internal(x, optimize=optimize)
  return ops.IndexedSlices(x, range(0, x_shape[0]), x_shape) 

def _as_indexed_slices(x, optimize=True):
  """Convert 'x' to IndexedSlices.

  Convert a dense Tensor to a block-sparse IndexedSlices.

  Args:
    x: Either a Tensor object, or an IndexedSlices object.
    optimize: if true, attempt to optimize the conversion of 'x'.

  Returns:
    An IndexedSlices object.

  Raises:
    TypeError: If 'x' is not a Tensor or an IndexedSlices object.
  """
  # TODO(touts): op_scope
  if not isinstance(x, (ops.Tensor, ops.IndexedSlices)):
    raise TypeError("Not a Tensor or IndexedSlices: %s" % type(x))
  if isinstance(x, ops.IndexedSlices):
    return x
  x_shape = array_ops.shape_internal(x, optimize=optimize)
  return ops.IndexedSlices(x, range(0, x_shape[0]), x_shape) 

def _as_indexed_slices(x, optimize=True):
  """Convert 'x' to IndexedSlices.

  Convert a dense Tensor to a block-sparse IndexedSlices.

  Args:
    x: Either a Tensor object, or an IndexedSlices object.
    optimize: if true, attempt to optimize the conversion of 'x'.

  Returns:
    An IndexedSlices object.

  Raises:
    TypeError: If 'x' is not a Tensor or an IndexedSlices object.
  """
  # TODO(touts): op_scope
  if not isinstance(x, (ops.Tensor, ops.IndexedSlices)):
    raise TypeError("Not a Tensor or IndexedSlices: %s" % type(x))
  if isinstance(x, ops.IndexedSlices):
    return x
  x_shape = array_ops.shape_internal(x, optimize=optimize)
  return ops.IndexedSlices(x, range(0, x_shape[0]), x_shape) 

def _as_indexed_slices(x, optimize=True):
  """Convert 'x' to IndexedSlices.

  Convert a dense Tensor to a block-sparse IndexedSlices.

  Args:
    x: Either a Tensor object, or an IndexedSlices object.
    optimize: if true, attempt to optimize the conversion of 'x'.

  Returns:
    An IndexedSlices object.

  Raises:
    TypeError: If 'x' is not a Tensor or an IndexedSlices object.
  """
  # TODO(touts): op_scope
  if not isinstance(x, (ops.Tensor, ops.IndexedSlices)):
    raise TypeError("Not a Tensor or IndexedSlices: %s" % type(x))
  if isinstance(x, ops.IndexedSlices):
    return x
  x_shape = array_ops.shape_internal(x, optimize=optimize)
  return ops.IndexedSlices(x, range(0, x_shape[0]), x_shape) 

def ZerosLikeOutsideLoop(op, index):
  """Create zeros_like for the specified output of an op."""
  val = op.outputs[index]
  if not IsSwitch(op):
    return array_ops.zeros_like(val, optimize=False)
  else:
    op_ctxt = op._get_control_flow_context()
    pred = op_ctxt.pred
    branch = op_ctxt.branch
    switch_val = switch(op.inputs[0], pred)[1 - branch]
    zeros_shape = array_ops.shape_internal(switch_val, optimize=False)
    return array_ops.zeros(zeros_shape, dtype=val.dtype) 

def PostProcessing(self):
    """Perform postprocessing at the end of gradients().

    We have created the gradient graph at this point. So this function
    can be used to perform any postprocessing on the gradient graph.
    We currently perform the following postprocessing:
      1. Patch the gradient graph if the output of a loop variable
         doesn't depend on its input.
    """
    for _, grad_state in self._map.items():
      for _, b_merge in grad_state.switch_map.items():
        if b_merge.op.inputs[0] == b_merge.op.inputs[1]:
          # The value of this loop variable at iteration i+1 doesn't
          # depend on its value at iteration i. So use zeros as the
          # gradients for all iterations > 0.
          dtype = b_merge.op.inputs[0].dtype
          shape = b_merge.op.inputs[0].get_shape()
          # pylint: disable=protected-access
          if shape.is_fully_defined():
            grad_state.grad_context.Enter()
            # Create a zeros and use it for iterations > 0.
            grad_val = constant_op.constant(0, dtype=dtype, shape=shape)
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          else:
            # Create a zeros in the outer grad context.
            outer_grad_ctxt = grad_state.grad_context.outer_context
            if outer_grad_ctxt: outer_grad_ctxt.Enter()
            enter_grad_op = b_merge.op.inputs[0].op
            enter_grad = enter_grad_op.inputs[0]
            grad_shape = array_ops.shape_internal(enter_grad, optimize=False)
            grad_val = array_ops.zeros(grad_shape)
            if outer_grad_ctxt: outer_grad_ctxt.Exit()
            # Use the zeros for iterations > 0.
            grad_state.grad_context.Enter()
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          b_merge.op._update_input(1, next_grad_val)
          # pylint: enable=protected-access 

def ZerosLikeOutsideLoop(op, index):
  """Create zeros_like for the specified output of an op."""
  val = op.outputs[index]
  if not IsSwitch(op):
    return array_ops.zeros_like(val, optimize=False)
  else:
    op_ctxt = op._get_control_flow_context()
    pred = op_ctxt.pred
    branch = op_ctxt.branch
    switch_val = switch(op.inputs[0], pred)[1 - branch]
    zeros_shape = array_ops.shape_internal(switch_val, optimize=False)
    return array_ops.zeros(zeros_shape, dtype=val.dtype) 

def PostProcessing(self):
    """Perform postprocessing at the end of gradients().

    We have created the gradient graph at this point. So this function
    can be used to perform any postprocessing on the gradient graph.
    We currently perform the following postprocessing:
      1. Patch the gradient graph if the output of a loop variable
         doesn't depend on its input.
    """
    for _, grad_state in self._map.items():
      for _, b_merge in grad_state.switch_map.items():
        if b_merge.op.inputs[0] == b_merge.op.inputs[1]:
          # The value of this loop variable at iteration i+1 doesn't
          # depend on its value at iteration i. So use zeros as the
          # gradients for all iterations > 0.
          dtype = b_merge.op.inputs[0].dtype
          shape = b_merge.op.inputs[0].get_shape()
          # pylint: disable=protected-access
          if shape.is_fully_defined():
            grad_state.grad_context.Enter()
            # Create a zeros and use it for iterations > 0.
            grad_val = constant_op.constant(0, dtype=dtype, shape=shape)
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          else:
            # Create a zeros in the outer grad context.
            outer_grad_ctxt = grad_state.grad_context.outer_context
            if outer_grad_ctxt: outer_grad_ctxt.Enter()
            enter_grad_op = b_merge.op.inputs[0].op
            enter_grad = enter_grad_op.inputs[0]
            grad_shape = array_ops.shape_internal(enter_grad, optimize=False)
            grad_val = array_ops.zeros(grad_shape)
            if outer_grad_ctxt: outer_grad_ctxt.Exit()
            # Use the zeros for iterations > 0.
            grad_state.grad_context.Enter()
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          b_merge.op._update_input(1, next_grad_val)
          # pylint: enable=protected-access 

def PostProcessing(self):
    """Perform postprocessing at the end of gradients().

    We have created the gradient graph at this point. So this function
    can be used to perform any postprocessing on the gradient graph.
    We currently perform the following postprocessing:
      1. Patch the gradient graph if the output of a loop variable
         doesn't depend on its input.
    """
    for _, grad_state in self._map.items():
      for _, b_merge in grad_state.switch_map.items():
        if b_merge.op.inputs[0] == b_merge.op.inputs[1]:
          # The value of this loop variable at iteration i+1 doesn't
          # depend on its value at iteration i. So use zeros as the
          # gradients for all iterations > 0.
          dtype = b_merge.op.inputs[0].dtype
          shape = b_merge.op.inputs[0].get_shape()
          # pylint: disable=protected-access
          if shape.is_fully_defined():
            grad_state.grad_context.Enter()
            # Create a zeros and use it for iterations > 0.
            grad_val = constant_op.constant(0, dtype=dtype, shape=shape)
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          else:
            # Create a zeros in the outer grad context.
            outer_grad_ctxt = grad_state.grad_context.outer_context
            if outer_grad_ctxt: outer_grad_ctxt.Enter()
            enter_grad_op = b_merge.op.inputs[0].op
            enter_grad = enter_grad_op.inputs[0]
            grad_shape = array_ops.shape_internal(enter_grad, optimize=False)
            grad_val = array_ops.zeros(grad_shape)
            if outer_grad_ctxt: outer_grad_ctxt.Exit()
            # Use the zeros for iterations > 0.
            grad_state.grad_context.Enter()
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          b_merge.op._update_input(1, next_grad_val)
          # pylint: enable=protected-access 

def PostProcessing(self):
    """Perform postprocessing at the end of gradients().

    We have created the gradient graph at this point. So this function
    can be used to perform any postprocessing on the gradient graph.
    We currently perform the following postprocessing:
      1. Patch the gradient graph if the output of a loop variable
         doesn't depend on its input.
    """
    for _, grad_state in self._map.items():
      for _, b_merge in grad_state.switch_map.items():
        if b_merge.op.inputs[0] == b_merge.op.inputs[1]:
          # The value of this loop variable at iteration i+1 doesn't
          # depend on its value at iteration i. So use zeros as the
          # gradients for all iterations > 0.
          dtype = b_merge.op.inputs[0].dtype
          shape = b_merge.op.inputs[0].get_shape()
          # pylint: disable=protected-access
          if shape.is_fully_defined():
            grad_state.grad_context.Enter()
            # Create a zeros and use it for iterations > 0.
            grad_val = constant_op.constant(0, dtype=dtype, shape=shape)
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          else:
            # Create a zeros in the outer grad context.
            outer_grad_ctxt = grad_state.grad_context.outer_context
            if outer_grad_ctxt: outer_grad_ctxt.Enter()
            enter_grad_op = b_merge.op.inputs[0].op
            enter_grad = enter_grad_op.inputs[0]
            grad_shape = array_ops.shape_internal(enter_grad, optimize=False)
            grad_val = array_ops.zeros(grad_shape)
            if outer_grad_ctxt: outer_grad_ctxt.Exit()
            # Use the zeros for iterations > 0.
            grad_state.grad_context.Enter()
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          b_merge.op._update_input(1, next_grad_val)
          # pylint: enable=protected-access 

def PostProcessing(self):
    """Perform postprocessing at the end of gradients().

    We have created the gradient graph at this point. So this function
    can be used to perform any postprocessing on the gradient graph.
    We currently perform the following postprocessing:
      1. Patch the gradient graph if the output of a loop variable
         doesn't depend on its input.
    """
    for _, grad_state in self._map.items():
      for _, b_merge in grad_state.switch_map.items():
        if b_merge.op.inputs[0] == b_merge.op.inputs[1]:
          # The value of this loop variable at iteration i+1 doesn't
          # depend on its value at iteration i. So use zeros as the
          # gradients for all iterations > 0.
          dtype = b_merge.op.inputs[0].dtype
          shape = b_merge.op.inputs[0].get_shape()
          # pylint: disable=protected-access
          if shape.is_fully_defined():
            grad_state.grad_context.Enter()
            # Create a zeros and use it for iterations > 0.
            grad_val = constant_op.constant(0, dtype=dtype, shape=shape)
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          else:
            # Create a zeros in the outer grad context.
            outer_grad_ctxt = grad_state.grad_context.outer_context
            if outer_grad_ctxt: outer_grad_ctxt.Enter()
            enter_grad_op = b_merge.op.inputs[0].op
            enter_grad = enter_grad_op.inputs[0]
            grad_shape = array_ops.shape_internal(enter_grad, optimize=False)
            grad_val = array_ops.zeros(grad_shape)
            if outer_grad_ctxt: outer_grad_ctxt.Exit()
            # Use the zeros for iterations > 0.
            grad_state.grad_context.Enter()
            next_grad_val = _NextIteration(grad_val)
            grad_state.grad_context.Exit()
          b_merge.op._update_input(1, next_grad_val)
          # pylint: enable=protected-access 

def ZerosLikeOutsideLoop(op, index):
  """Create zeros_like for the specified output of an op."""
  val = op.outputs[index]
  if not IsSwitch(op):
    return array_ops.zeros_like(val, optimize=False)
  else:
    op_ctxt = op._get_control_flow_context()
    pred = op_ctxt.pred
    branch = op_ctxt.branch
    switch_val = switch(op.inputs[0], pred)[1 - branch]
    zeros_shape = array_ops.shape_internal(switch_val, optimize=False)
    return array_ops.zeros(zeros_shape, dtype=val.dtype) 

def ZerosLike(self, op, index):
    """Create zeros_like for the specified output of an op.

    If op is in a while loop that is part of gradients(), this method
    must be called in its grad loop context.

    Args:
      op: A tensorflow operation.
      index: the index for a specific output of the op.

    Returns:
      A zero tensor of the same shape of op.outputs[index].
    """
    if IsLoopSwitch(op): return None
    dead_branch = IsSwitch(op)
    forward_ctxt = _GetWhileContext(op)
    grad_state = self._map.get(forward_ctxt)
    if grad_state is None:
      # op is not in a while loop that is part of gradients().
      return ZerosLikeOutsideLoop(op, index)
    op_ctxt = op._get_control_flow_context()
    val = ops.convert_to_tensor(op.outputs[index], name="tensor")
    shape = val.get_shape()
    if shape.is_fully_defined():
      # If the shape is known statically, just create a zero tensor with
      # the right shape in the grad loop context.
      result = constant_op.constant(0, shape=shape.dims, dtype=val.dtype)
      if dead_branch:
        # op is a cond switch. Guard the zero tensor with a switch.
        pred = grad_state.history_map.get(op_ctxt.pred.name)
        branch = op_ctxt.branch
        result = _SwitchRefOrTensor(result, pred)[1 - branch]
    else:
      # Unknown shape so keep a history of the shape at runtime.
      if dead_branch:
        # Need to add a special switch to guard the value.
        pred = op_ctxt.pred
        branch = op_ctxt.branch
        op_ctxt.outer_context.Enter()
        val = _SwitchRefOrTensor(op.inputs[0], pred)[1 - branch]
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.outer_context.Exit()
        val.op._set_control_flow_context(op_ctxt)
        zeros_shape.op._set_control_flow_context(op_ctxt)
      else:
        op_ctxt.Enter()
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.Exit()

      # Add forward accumulator for shape.
      grad_state.grad_context.Exit()
      history_zeros_shape = grad_state.AddForwardAccumulator(
          zeros_shape, dead_branch=dead_branch)
      grad_state.grad_context.Enter()

      # Create a zero tensor with the right shape.
      shape = grad_state.AddBackPropAccumulatedValue(
          history_zeros_shape, zeros_shape, dead_branch)
      result = array_ops.zeros(shape, val.dtype)
    return result 

def ZerosLikeForExit(self, val):
    """Create zeros_like gradient for a loop exit.

    If the result of a loop variable is not used but is involved in
    computing the result of some needed loop variable, we create a
    zero-valued tensor that is fed as gradient for the Exit node of that
    loop variable. Note that val.op is an Exit, and this method must be
    called in the control flow context where gradients() is called.

    Args:
      val: The output tensor of an Exit op.

    Returns:
      A zero tensor of the same shape of val.
    """
    val_shape = val.get_shape()
    forward_ctxt = val.op._get_control_flow_context()
    outer_forward_ctxt = forward_ctxt.outer_context
    if outer_forward_ctxt:
      outer_forward_ctxt = outer_forward_ctxt.GetWhileContext()
    outer_grad_state = None
    if outer_forward_ctxt:
      outer_grad_state = self._map.get(outer_forward_ctxt)
    if outer_grad_state:
      # This is a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape in the right context.
        outer_grad_state.grad_context.Enter()
        result = array_ops.zeros(val_shape.dims, val.dtype)
        outer_grad_state.grad_context.Exit()
      else:
        # Only the shape of value is needed for backprop.
        forward_ctxt.outer_context.Enter()
        shape = array_ops.shape_internal(val, optimize=False)
        forward_ctxt.outer_context.Exit()
        # Save the shape to a stack.
        history_shape = outer_grad_state.AddForwardAccumulator(shape)
        # Get the shape back from the stack.
        outer_grad_ctxt = outer_grad_state.grad_context
        outer_grad_ctxt.Enter()
        real_shape = outer_grad_state.AddBackPropAccumulatedValue(
            history_shape, shape)
        result = array_ops.zeros(real_shape, val.dtype)
        outer_grad_ctxt.Exit()
    else:
      # This is not a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape.
        result = array_ops.zeros(val_shape.dims, val.dtype)
      else:
        result = array_ops.zeros_like(val, optimize=False)
    return result 

def ZerosLike(self, op, index):
    """Create zeros_like for the specified output of an op.

    If op is in a while loop that is part of gradients(), this method
    must be called in its grad loop context.

    Args:
      op: A tensorflow operation.
      index: the index for a specific output of the op.

    Returns:
      A zero tensor of the same shape of op.outputs[index].
    """
    if IsLoopSwitch(op): return None
    dead_branch = IsSwitch(op)
    forward_ctxt = _GetWhileContext(op)
    grad_state = self._map.get(forward_ctxt)
    if grad_state is None:
      # op is not in a while loop that is part of gradients().
      return ZerosLikeOutsideLoop(op, index)
    op_ctxt = op._get_control_flow_context()
    val = ops.convert_to_tensor(op.outputs[index], name="tensor")
    shape = val.get_shape()
    if shape.is_fully_defined():
      # If the shape is known statically, just create a zero tensor with
      # the right shape in the grad loop context.
      result = constant_op.constant(0, shape=shape.dims, dtype=val.dtype)
      if dead_branch:
        # op is a cond switch. Guard the zero tensor with a switch.
        pred = grad_state.history_map.get(op_ctxt.pred.name)
        branch = op_ctxt.branch
        result = _SwitchRefOrTensor(result, pred)[1 - branch]
    else:
      # Unknown shape so keep a history of the shape at runtime.
      if dead_branch:
        # Need to add a special switch to guard the value.
        pred = op_ctxt.pred
        branch = op_ctxt.branch
        op_ctxt.outer_context.Enter()
        val = _SwitchRefOrTensor(op.inputs[0], pred)[1 - branch]
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.outer_context.Exit()
        val.op._set_control_flow_context(op_ctxt)
        zeros_shape.op._set_control_flow_context(op_ctxt)
      else:
        op_ctxt.Enter()
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.Exit()

      # Add forward accumulator for shape.
      grad_state.grad_context.Exit()
      history_zeros_shape = grad_state.AddForwardAccumulator(
          zeros_shape, dead_branch=dead_branch)
      grad_state.grad_context.Enter()

      # Create a zero tensor with the right shape.
      shape = grad_state.AddBackPropAccumulatedValue(
          history_zeros_shape, zeros_shape, dead_branch)
      result = array_ops.zeros(shape, val.dtype)
    return result 

def ZerosLike(self, op, index):
    """Create zeros_like for the specified output of an op.

    If op is in a while loop that is part of gradients(), this method
    must be called in its grad loop context.

    Args:
      op: A tensorflow operation.
      index: the index for a specific output of the op.

    Returns:
      A zero tensor of the same shape of op.outputs[index].
    """
    if IsLoopSwitch(op): return None
    dead_branch = IsSwitch(op)
    forward_ctxt = _GetWhileContext(op)
    grad_state = self._map.get(forward_ctxt)
    if grad_state is None:
      # op is not in a while loop that is part of gradients().
      return ZerosLikeOutsideLoop(op, index)
    op_ctxt = op._get_control_flow_context()
    val = ops.convert_to_tensor(op.outputs[index], name="tensor")
    shape = val.get_shape()
    if shape.is_fully_defined():
      # If the shape is known statically, just create a zero tensor with
      # the right shape in the grad loop context.
      result = constant_op.constant(0, shape=shape.dims, dtype=val.dtype)
      if dead_branch:
        # op is a cond switch. Guard the zero tensor with a switch.
        pred = grad_state.history_map.get(op_ctxt.pred.name)
        branch = op_ctxt.branch
        result = _SwitchRefOrTensor(result, pred)[1 - branch]
    else:
      # Unknown shape so keep a history of the shape at runtime.
      if dead_branch:
        # Need to add a special switch to guard the value.
        pred = op_ctxt.pred
        branch = op_ctxt.branch
        op_ctxt.outer_context.Enter()
        val = _SwitchRefOrTensor(op.inputs[0], pred)[1 - branch]
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.outer_context.Exit()
        val.op._set_control_flow_context(op_ctxt)
        zeros_shape.op._set_control_flow_context(op_ctxt)
      else:
        op_ctxt.Enter()
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.Exit()

      # Add forward accumulator for shape.
      grad_state.grad_context.Exit()
      history_zeros_shape = grad_state.AddForwardAccumulator(
          zeros_shape, dead_branch=dead_branch)
      grad_state.grad_context.Enter()

      # Create a zero tensor with the right shape.
      shape = grad_state.AddBackpropAccumulatedValue(
          history_zeros_shape, zeros_shape, dead_branch)
      result = array_ops.zeros(shape, val.dtype)
    return result 

def ZerosLikeForExit(self, val):
    """Create zeros_like gradient for a loop exit.

    If the result of a loop variable is not used but is involved in
    computing the result of some needed loop variable, we create a
    zero-valued tensor that is fed as gradient for the Exit node of that
    loop variable. Note that val.op is an Exit, and this method must be
    called in the control flow context where gradients() is called.

    Args:
      val: The output tensor of an Exit op.

    Returns:
      A zero tensor of the same shape of val.
    """
    val_shape = val.get_shape()
    forward_ctxt = val.op._get_control_flow_context()
    outer_forward_ctxt = forward_ctxt.outer_context
    if outer_forward_ctxt:
      outer_forward_ctxt = outer_forward_ctxt.GetWhileContext()
    outer_grad_state = None
    if outer_forward_ctxt:
      outer_grad_state = self._map.get(outer_forward_ctxt)
    if outer_grad_state:
      # This is a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape in the right context.
        outer_grad_state.grad_context.Enter()
        result = array_ops.zeros(val_shape.dims, val.dtype)
        outer_grad_state.grad_context.Exit()
      else:
        # Only the shape of value is needed for backprop.
        forward_ctxt.outer_context.Enter()
        shape = array_ops.shape_internal(val, optimize=False)
        forward_ctxt.outer_context.Exit()
        # Save the shape to a stack.
        history_shape = outer_grad_state.AddForwardAccumulator(shape)
        # Get the shape back from the stack.
        outer_grad_ctxt = outer_grad_state.grad_context
        outer_grad_ctxt.Enter()
        real_shape = outer_grad_state.AddBackpropAccumulatedValue(
            history_shape, shape)
        result = array_ops.zeros(real_shape, val.dtype)
        outer_grad_ctxt.Exit()
    else:
      # This is not a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape.
        result = array_ops.zeros(val_shape.dims, val.dtype)
      else:
        result = array_ops.zeros_like(val, optimize=False)
    return result 

def ZerosLikeForExit(self, val):
    """Create zeros_like gradient for a loop exit.

    If the result of a loop variable is not used but is involved in
    computing the result of some needed loop variable, we create a
    zero-valued tensor that is fed as gradient for the Exit node of that
    loop variable. Note that val.op is an Exit, and this method must be
    called in the control flow context where gradients() is called.

    Args:
      val: The output tensor of an Exit op.

    Returns:
      A zero tensor of the same shape of val.
    """
    val_shape = val.get_shape()
    forward_ctxt = val.op._get_control_flow_context()
    outer_forward_ctxt = forward_ctxt.outer_context
    if outer_forward_ctxt:
      outer_forward_ctxt = outer_forward_ctxt.GetWhileContext()
    outer_grad_state = None
    if outer_forward_ctxt:
      outer_grad_state = self._map.get(outer_forward_ctxt)
    if outer_grad_state:
      # This is a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape in the right context.
        outer_grad_state.grad_context.Enter()
        result = array_ops.zeros(val_shape.dims, val.dtype)
        outer_grad_state.grad_context.Exit()
      else:
        # Only the shape of value is needed for backprop.
        forward_ctxt.outer_context.Enter()
        shape = array_ops.shape_internal(val, optimize=False)
        forward_ctxt.outer_context.Exit()
        # Save the shape to a stack.
        history_shape = outer_grad_state.AddForwardAccumulator(shape)
        # Get the shape back from the stack.
        outer_grad_ctxt = outer_grad_state.grad_context
        outer_grad_ctxt.Enter()
        real_shape = outer_grad_state.AddBackPropAccumulatedValue(
            history_shape, shape)
        result = array_ops.zeros(real_shape, val.dtype)
        outer_grad_ctxt.Exit()
    else:
      # This is not a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape.
        result = array_ops.zeros(val_shape.dims, val.dtype)
      else:
        result = array_ops.zeros_like(val, optimize=False)
    return result 

def ZerosLike(self, op, index):
    """Create zeros_like for the specified output of an op.

    If op is in a while loop that is part of gradients(), this method
    must be called in its grad loop context.

    Args:
      op: A tensorflow operation.
      index: the index for a specific output of the op.

    Returns:
      A zero tensor of the same shape of op.outputs[index].
    """
    if IsLoopSwitch(op): return None
    dead_branch = IsSwitch(op)
    forward_ctxt = _GetWhileContext(op)
    grad_state = self._map.get(forward_ctxt)
    if grad_state is None:
      # op is not in a while loop that is part of gradients().
      return ZerosLikeOutsideLoop(op, index)
    op_ctxt = op._get_control_flow_context()
    val = ops.convert_to_tensor(op.outputs[index], name="tensor")
    shape = val.get_shape()
    if shape.is_fully_defined():
      # If the shape is known statically, just create a zero tensor with
      # the right shape in the grad loop context.
      result = constant_op.constant(0, shape=shape.dims, dtype=val.dtype)
      if dead_branch:
        # op is a cond switch. Guard the zero tensor with a switch.
        pred = grad_state.history_map.get(op_ctxt.pred.name)
        branch = op_ctxt.branch
        result = _SwitchRefOrTensor(result, pred)[1 - branch]
    else:
      # Unknown shape so keep a history of the shape at runtime.
      if dead_branch:
        # Need to add a special switch to guard the value.
        pred = op_ctxt.pred
        branch = op_ctxt.branch
        op_ctxt.outer_context.Enter()
        val = _SwitchRefOrTensor(op.inputs[0], pred)[1 - branch]
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.outer_context.Exit()
        val.op._set_control_flow_context(op_ctxt)
        zeros_shape.op._set_control_flow_context(op_ctxt)
      else:
        op_ctxt.Enter()
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.Exit()

      # Add forward accumulator for shape.
      grad_state.grad_context.Exit()
      history_zeros_shape = grad_state.AddForwardAccumulator(
          zeros_shape, dead_branch=dead_branch)
      grad_state.grad_context.Enter()

      # Create a zero tensor with the right shape.
      shape = grad_state.AddBackPropAccumulatedValue(
          history_zeros_shape, zeros_shape, dead_branch)
      result = array_ops.zeros(shape, val.dtype)
    return result 

def ZerosLikeForExit(self, val):
    """Create zeros_like gradient for a loop exit.

    If the result of a loop variable is not used but is involved in
    computing the result of some needed loop variable, we create a
    zero-valued tensor that is fed as gradient for the Exit node of that
    loop variable. Note that val.op is an Exit, and this method must be
    called in the control flow context where gradients() is called.

    Args:
      val: The output tensor of an Exit op.

    Returns:
      A zero tensor of the same shape of val.
    """
    val_shape = val.get_shape()
    forward_ctxt = val.op._get_control_flow_context()
    outer_forward_ctxt = forward_ctxt.outer_context
    if outer_forward_ctxt:
      outer_forward_ctxt = outer_forward_ctxt.GetWhileContext()
    outer_grad_state = None
    if outer_forward_ctxt:
      outer_grad_state = self._map.get(outer_forward_ctxt)
    if outer_grad_state:
      # This is a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape in the right context.
        outer_grad_state.grad_context.Enter()
        result = array_ops.zeros(val_shape.dims, val.dtype)
        outer_grad_state.grad_context.Exit()
      else:
        # Only the shape of value is needed for backprop.
        forward_ctxt.outer_context.Enter()
        shape = array_ops.shape_internal(val, optimize=False)
        forward_ctxt.outer_context.Exit()
        # Save the shape to a stack.
        history_shape = outer_grad_state.AddForwardAccumulator(shape)
        # Get the shape back from the stack.
        outer_grad_ctxt = outer_grad_state.grad_context
        outer_grad_ctxt.Enter()
        real_shape = outer_grad_state.AddBackPropAccumulatedValue(
            history_shape, shape)
        result = array_ops.zeros(real_shape, val.dtype)
        outer_grad_ctxt.Exit()
    else:
      # This is not a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape.
        result = array_ops.zeros(val_shape.dims, val.dtype)
      else:
        result = array_ops.zeros_like(val, optimize=False)
    return result 

def ZerosLike(self, op, index):
    """Create zeros_like for the specified output of an op.

    If op is in a while loop that is part of gradients(), this method
    must be called in its grad loop context.

    Args:
      op: A tensorflow operation.
      index: the index for a specific output of the op.

    Returns:
      A zero tensor of the same shape of op.outputs[index].
    """
    if IsLoopSwitch(op): return None
    dead_branch = IsSwitch(op)
    forward_ctxt = _GetWhileContext(op)
    grad_state = self._map.get(forward_ctxt)
    if grad_state is None:
      # op is not in a while loop that is part of gradients().
      return ZerosLikeOutsideLoop(op, index)
    op_ctxt = op._get_control_flow_context()
    val = ops.convert_to_tensor(op.outputs[index], name="tensor")
    shape = val.get_shape()
    if shape.is_fully_defined():
      # If the shape is known statically, just create a zero tensor with
      # the right shape in the grad loop context.
      result = constant_op.constant(0, shape=shape.dims, dtype=val.dtype)
      if dead_branch:
        # op is a cond switch. Guard the zero tensor with a switch.
        pred = grad_state.history_map.get(op_ctxt.pred.name)
        branch = op_ctxt.branch
        result = _SwitchRefOrTensor(result, pred)[1 - branch]
    else:
      # Unknown shape so keep a history of the shape at runtime.
      if dead_branch:
        # Need to add a special switch to guard the value.
        pred = op_ctxt.pred
        branch = op_ctxt.branch
        op_ctxt.outer_context.Enter()
        val = _SwitchRefOrTensor(op.inputs[0], pred)[1 - branch]
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.outer_context.Exit()
        val.op._set_control_flow_context(op_ctxt)
        zeros_shape.op._set_control_flow_context(op_ctxt)
      else:
        op_ctxt.Enter()
        zeros_shape = array_ops.shape_internal(val, optimize=False)
        op_ctxt.Exit()

      # Add forward accumulator for shape.
      grad_state.grad_context.Exit()
      history_zeros_shape = grad_state.AddForwardAccumulator(
          zeros_shape, dead_branch=dead_branch)
      grad_state.grad_context.Enter()

      # Create a zero tensor with the right shape.
      shape = grad_state.AddBackPropAccumulatedValue(
          history_zeros_shape, zeros_shape, dead_branch)
      result = array_ops.zeros(shape, val.dtype)
    return result 

def ZerosLikeForExit(self, val):
    """Create zeros_like gradient for a loop exit.

    If the result of a loop variable is not used but is involved in
    computing the result of some needed loop variable, we create a
    zero-valued tensor that is fed as gradient for the Exit node of that
    loop variable. Note that val.op is an Exit, and this method must be
    called in the control flow context where gradients() is called.

    Args:
      val: The output tensor of an Exit op.

    Returns:
      A zero tensor of the same shape of val.
    """
    val_shape = val.get_shape()
    forward_ctxt = val.op._get_control_flow_context()
    outer_forward_ctxt = forward_ctxt.outer_context
    if outer_forward_ctxt:
      outer_forward_ctxt = outer_forward_ctxt.GetWhileContext()
    outer_grad_state = None
    if outer_forward_ctxt:
      outer_grad_state = self._map.get(outer_forward_ctxt)
    if outer_grad_state:
      # This is a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape in the right context.
        outer_grad_state.grad_context.Enter()
        result = array_ops.zeros(val_shape.dims, val.dtype)
        outer_grad_state.grad_context.Exit()
      else:
        # Only the shape of value is needed for backprop.
        forward_ctxt.outer_context.Enter()
        shape = array_ops.shape_internal(val, optimize=False)
        forward_ctxt.outer_context.Exit()
        # Save the shape to a stack.
        history_shape = outer_grad_state.AddForwardAccumulator(shape)
        # Get the shape back from the stack.
        outer_grad_ctxt = outer_grad_state.grad_context
        outer_grad_ctxt.Enter()
        real_shape = outer_grad_state.AddBackPropAccumulatedValue(
            history_shape, shape)
        result = array_ops.zeros(real_shape, val.dtype)
        outer_grad_ctxt.Exit()
    else:
      # This is not a nested loop.
      if val_shape.is_fully_defined():
        # If the shape is known statically, just create a zero tensor
        # with the right shape.
        result = array_ops.zeros(val_shape.dims, val.dtype)
      else:
        result = array_ops.zeros_like(val, optimize=False)
    return result