tensorflow:: ops:: SparseApplyCenteredRMSProp

#include <training_ops.h>

Update '*var' according to the centered RMSProp algorithm.

Summary

The centered RMSProp algorithm uses an estimate of the centered second moment (i.e., the variance) for normalization, as opposed to regular RMSProp, which uses the (uncentered) second moment. This often helps with training, but is slightly more expensive in terms of computation and memory.

Note that in dense implementation of this algorithm, mg, ms, and mom will update even if the grad is zero, but in this sparse implementation, mg, ms, and mom will not update in iterations during which the grad is zero.

mean_square = decay * mean_square + (1-decay) * gradient ** 2 mean_grad = decay * mean_grad + (1-decay) * gradient Delta = learning_rate * gradient / sqrt(mean_square + epsilon - mean_grad ** 2)

$$ms <- rho * ms_{t-1} + (1-rho) * grad * grad$$
$$mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)$$
$$var <- var - mom$$

Args:

  • scope: A Scope object
  • var: Should be from a Variable().
  • mg: Should be from a Variable().
  • ms: Should be from a Variable().
  • mom: Should be from a Variable().
  • lr: Scaling factor. Must be a scalar.
  • rho: Decay rate. Must be a scalar.
  • epsilon: Ridge term. Must be a scalar.
  • grad: The gradient.
  • indices: A vector of indices into the first dimension of var, ms and mom.

Optional attributes (see Attrs ):

  • use_locking: If True , updating of the var, mg, ms, and mom tensors is protected by a lock; otherwise the behavior is undefined, but may exhibit less contention.

Returns:

Constructors and Destructors

SparseApplyCenteredRMSProp (const :: tensorflow::Scope & scope, :: tensorflow::Input var, :: tensorflow::Input mg, :: tensorflow::Input ms, :: tensorflow::Input mom, :: tensorflow::Input lr, :: tensorflow::Input rho, :: tensorflow::Input momentum, :: tensorflow::Input epsilon, :: tensorflow::Input grad, :: tensorflow::Input indices)
SparseApplyCenteredRMSProp (const :: tensorflow::Scope & scope, :: tensorflow::Input var, :: tensorflow::Input mg, :: tensorflow::Input ms, :: tensorflow::Input mom, :: tensorflow::Input lr, :: tensorflow::Input rho, :: tensorflow::Input momentum, :: tensorflow::Input epsilon, :: tensorflow::Input grad, :: tensorflow::Input indices, const SparseApplyCenteredRMSProp::Attrs & attrs)

Public attributes

operation
out

Public functions

node () const
::tensorflow::Node *
operator::tensorflow::Input () const
operator::tensorflow::Output () const

Public static functions

UseLocking (bool x)

Structs

tensorflow:: ops:: SparseApplyCenteredRMSProp:: Attrs

Optional attribute setters for SparseApplyCenteredRMSProp .

Public attributes

operation 

Operation operation

out 

::tensorflow::Output out

Public functions

SparseApplyCenteredRMSProp 

 SparseApplyCenteredRMSProp(
  const ::tensorflow::Scope & scope,
  ::tensorflow::Input var,
  ::tensorflow::Input mg,
  ::tensorflow::Input ms,
  ::tensorflow::Input mom,
  ::tensorflow::Input lr,
  ::tensorflow::Input rho,
  ::tensorflow::Input momentum,
  ::tensorflow::Input epsilon,
  ::tensorflow::Input grad,
  ::tensorflow::Input indices
)

SparseApplyCenteredRMSProp 

 SparseApplyCenteredRMSProp(
  const ::tensorflow::Scope & scope,
  ::tensorflow::Input var,
  ::tensorflow::Input mg,
  ::tensorflow::Input ms,
  ::tensorflow::Input mom,
  ::tensorflow::Input lr,
  ::tensorflow::Input rho,
  ::tensorflow::Input momentum,
  ::tensorflow::Input epsilon,
  ::tensorflow::Input grad,
  ::tensorflow::Input indices,
  const SparseApplyCenteredRMSProp::Attrs & attrs
)

node 

::tensorflow::Node * node() const

operator::tensorflow::Input 

 operator::tensorflow::Input() const

operator::tensorflow::Output 

 operator::tensorflow::Output() const

Public static functions

UseLocking 

Attrs UseLocking(
  bool x
)