Update '*var' according to the centered RMSProp algorithm.
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)
$$ms <- rho * ms_{t-1} + (1-rho) * grad * grad$$ $$mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)$$ $$var <- var - mom$$
Nested Classes
class | SparseApplyCenteredRmsProp.Options | Optional attributes for SparseApplyCenteredRmsProp
|
Constants
String | OP_NAME | The name of this op, as known by TensorFlow core engine |
Public Methods
Output<T> |
asOutput()
Returns the symbolic handle of the tensor.
|
static <T extends TType> SparseApplyCenteredRmsProp<T> |
create(Scope scope, Operand<T> var, Operand<T> mg, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, Options... options)
Factory method to create a class wrapping a new SparseApplyCenteredRmsProp operation.
|
Output<T> |
out()
Same as "var".
|
static SparseApplyCenteredRmsProp.Options |
useLocking(Boolean useLocking)
|
Inherited Methods
Constants
public static final String OP_NAME
The name of this op, as known by TensorFlow core engine
Public Methods
public Output<T> asOutput ()
Returns the symbolic handle of the tensor.
Inputs to TensorFlow operations are outputs of another TensorFlow operation. This method is used to obtain a symbolic handle that represents the computation of the input.
public static SparseApplyCenteredRmsProp<T> create (Scope scope, Operand<T> var, Operand<T> mg, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, Options... options)
Factory method to create a class wrapping a new SparseApplyCenteredRmsProp operation.
Parameters
scope | current scope |
---|---|
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. |
options | carries optional attributes values |
Returns
- a new instance of SparseApplyCenteredRmsProp
public static SparseApplyCenteredRmsProp.Options useLocking (Boolean useLocking)
Parameters
useLocking | 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. |
---|