Class TowerOptimizer
Inherits From: Optimizer
Gathers gradients from all towers and reduces them in the last one.
__init__
__init__(optimizer_or_optimizer_fn)
Wrap an existing optimizer for gathering gradients across towers. (deprecated)
Each invocation of model_fn has to call the same optimizers in the same order.
Multiple optimizers that use the same or different losses are supported.
If TowerOptimizer is used but replicate_model_fn
isn't, then no
aggregation will happen. All calls will simply be forwarded to the
underlying optimizer. The behavior is similar if there is only one tower.
If TowerOptimizer is used together with SyncReplicasOptimizer that wraps the user's optimizer, then it's the SyncReplicasOptimizer that needs to be wrapped with TowerOptimizer.
Args:
optimizer_or_optimizer_fn
: an instance of optimizer to wrap. That instance is going to be used for optimizer-specific logic. This can also be a no-argument function that returns such an optimizer instance.
Methods
tf.contrib.estimator.TowerOptimizer.apply_gradients
apply_gradients(
grads_and_vars,
global_step=None,
**kwargs
)
Collect gradients updates to apply them with the last tower.
tf.contrib.estimator.TowerOptimizer.compute_gradients
compute_gradients(
loss,
*args,
**kwargs
)
Compute gradients, but first, if needed, scale the loss.
tf.contrib.estimator.TowerOptimizer.get_name
get_name(
*args,
**kwargs
)
tf.contrib.estimator.TowerOptimizer.get_slot
get_slot(
*args,
**kwargs
)
Return a slot named name
created for var
by the Optimizer.
Some Optimizer
subclasses use additional variables. For example
Momentum
and Adagrad
use variables to accumulate updates. This method
gives access to these Variable
objects if for some reason you need them.
Use get_slot_names()
to get the list of slot names created by the
Optimizer
.
Args:
var
: A variable passed tominimize()
orapply_gradients()
.name
: A string.
Returns:
The Variable
for the slot if it was created, None
otherwise.
tf.contrib.estimator.TowerOptimizer.get_slot_names
get_slot_names(
*args,
**kwargs
)
Return a list of the names of slots created by the Optimizer
.
See get_slot()
.
Returns:
A list of strings.
tf.contrib.estimator.TowerOptimizer.has_been_used
@staticmethod
has_been_used()
tf.contrib.estimator.TowerOptimizer.minimize
minimize(
loss,
global_step=None,
var_list=None,
gate_gradients=GATE_OP,
aggregation_method=None,
colocate_gradients_with_ops=False,
name=None,
grad_loss=None
)
Add operations to minimize loss
by updating var_list
.
This method simply combines calls compute_gradients()
and
apply_gradients()
. If you want to process the gradient before applying
them call compute_gradients()
and apply_gradients()
explicitly instead
of using this function.
Args:
loss
: ATensor
containing the value to minimize.global_step
: OptionalVariable
to increment by one after the variables have been updated.var_list
: Optional list or tuple ofVariable
objects to update to minimizeloss
. Defaults to the list of variables collected in the graph under the keyGraphKeys.TRAINABLE_VARIABLES
.gate_gradients
: How to gate the computation of gradients. Can beGATE_NONE
,GATE_OP
, orGATE_GRAPH
.aggregation_method
: Specifies the method used to combine gradient terms. Valid values are defined in the classAggregationMethod
.colocate_gradients_with_ops
: If True, try colocating gradients with the corresponding op.name
: Optional name for the returned operation.grad_loss
: Optional. ATensor
holding the gradient computed forloss
.
Returns:
An Operation that updates the variables in var_list
. If global_step
was not None
, that operation also increments global_step
.
Raises:
ValueError
: If some of the variables are notVariable
objects.
Eager Compatibility
When eager execution is enabled, loss
should be a Python function that
takes no arguments and computes the value to be minimized. Minimization (and
gradient computation) is done with respect to the elements of var_list
if
not None, else with respect to any trainable variables created during the
execution of the loss
function. gate_gradients
, aggregation_method
,
colocate_gradients_with_ops
and grad_loss
are ignored when eager
execution is enabled.
tf.contrib.estimator.TowerOptimizer.variables
variables(
*args,
**kwargs
)
A list of variables which encode the current state of Optimizer
.
Includes slot variables and additional global variables created by the optimizer in the current default graph.
Returns:
A list of variables.