def assign_sub(self, delta, use_locking=None, name=None, read_value=True):
"""Subtracts a value from this variable.
Args:
delta: A `Tensor`. The value to subtract from this variable.
use_locking: If `True`, use locking during the operation.
name: The name to use for the operation.
read_value: A `bool`. Whether to read and return the new value of the
variable or not.
Returns:
If `read_value` is `True`, this method will return the new value of the
variable after the assignment has completed. Otherwise, when in graph mode
it will return the `Operation` that does the assignment, and when in eager
mode it will return `None`.
"""
# TODO(apassos): this here and below is not atomic. Consider making it
# atomic if there's a way to do so without a performance cost for those who
# don't need it.
with _handle_graph(self.handle), self._assign_dependencies():
assign_sub_op = gen_resource_variable_ops.assign_sub_variable_op(
self.handle, ops.convert_to_tensor(delta, dtype=self.dtype),
name=name)
if read_value:
return self._lazy_read(assign_sub_op)
return assign_sub_op
def assign_sub(self, delta, use_locking=None, name=None):
# TODO(apassos): this here and below is not atomic. Consider making it
# atomic if there's a way to do so without a performance cost for those who
# don't need it.
return self._lazy_read(gen_resource_variable_ops.assign_sub_variable_op(
self.handle,
ops.convert_to_tensor(delta, dtype=self.dtype),
name=name))
def assign_sub(self, delta, use_locking=None, name=None, read_value=True):
del use_locking
with _handle_graph(self.handle), self._assign_dependencies():
assign_sub_op = gen_resource_variable_ops.assign_sub_variable_op(
self.handle,
ops.convert_to_tensor(delta, dtype=self.dtype),
name=name)
if read_value:
return self._read_variable_op()
return assign_sub_op
def _assign_moving_average(self, variable, value, one_minus_decay):
with ops.name_scope(None, 'AssignMovingAvg',
[variable, value, one_minus_decay]) as scope:
with ops.colocate_with(variable):
update_delta = (variable.read_value() - value) * one_minus_decay
if isinstance(variable, resource_variable_ops.ResourceVariable):
# state_ops.assign_sub does an extra read_variable_op after the
# assign. We avoid that here.
return gen_resource_variable_ops.assign_sub_variable_op(
variable.handle, update_delta, name=scope)
else:
return state_ops.assign_sub(variable, update_delta, name=scope)
def _assign_moving_average(self, variable, value, one_minus_decay):
with ops.name_scope(None, 'AssignMovingAvg',
[variable, value, one_minus_decay]) as scope:
with ops.colocate_with(variable):
update_delta = math_ops.multiply(
math_ops.subtract(variable.read_value(), value),
one_minus_decay)
if isinstance(variable,
resource_variable_ops.ResourceVariable):
# state_ops.assign_sub does an extra read_variable_op after the
# assign. We avoid that here.
return gen_resource_variable_ops.assign_sub_variable_op(
variable.handle, update_delta, name=scope)
else:
return state_ops.assign_sub(variable,
update_delta,
name=scope)
def assign_sub(self, delta, use_locking=None, name=None, read_value=True):
"""Subtracts a value from this variable.
Args:
delta: A `Tensor`. The value to subtract from this variable.
use_locking: If `True`, use locking during the operation.
name: The name to use for the operation.
read_value: A `bool`. Whether to read and return the new value of the
variable or not.
Returns:
If `read_value` is `True`, this method will return the new value of the
variable after the assignment has completed. Otherwise, when in graph mode
it will return the `Operation` that does the assignment, and when in eager
mode it will return `None`.
"""
# TODO(apassos): this here and below is not atomic. Consider making it
# atomic if there's a way to do so without a performance cost for those who
# don't need it.
assign_sub_op = gen_resource_variable_ops.assign_sub_variable_op(
self.handle, ops.convert_to_tensor(delta, dtype=self.dtype), name=name)
if read_value:
return self._lazy_read(assign_sub_op)
return assign_sub_op
