@Operator(group="train") public final class ResourceApplyCenteredRmsProp extends RawOp
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)
mg <- rho * mg_{t-1} + (1-rho) * grad ms <- rho * ms_{t-1} + (1-rho) * grad * grad mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms - mg * mg + epsilon) var <- var - mom
Modifier and Type | Class and Description |
---|---|
static class |
ResourceApplyCenteredRmsProp.Inputs<T extends TType> |
static class |
ResourceApplyCenteredRmsProp.Options
Optional attributes for
ResourceApplyCenteredRmsProp |
Modifier and Type | Field and Description |
---|---|
static String |
OP_NAME
The name of this op, as known by TensorFlow core engine
|
Constructor and Description |
---|
ResourceApplyCenteredRmsProp(Operation operation) |
Modifier and Type | Method and Description |
---|---|
static <T extends TType> |
create(Scope scope,
Operand<? extends TType> var,
Operand<? extends TType> mg,
Operand<? extends TType> ms,
Operand<? extends TType> mom,
Operand<T> lr,
Operand<T> rho,
Operand<T> momentum,
Operand<T> epsilon,
Operand<T> grad,
ResourceApplyCenteredRmsProp.Options... options)
Factory method to create a class wrapping a new ResourceApplyCenteredRMSProp operation.
|
static ResourceApplyCenteredRmsProp.Options |
useLocking(Boolean useLocking)
Sets the useLocking option.
|
public static final String OP_NAME
public ResourceApplyCenteredRmsProp(Operation operation)
@Endpoint(describeByClass=true) public static <T extends TType> ResourceApplyCenteredRmsProp create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> mg, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ResourceApplyCenteredRmsProp.Options... options)
T
- data type for ResourceApplyCenteredRMSProp
output and operandsscope
- current scopevar
- 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.momentum
- Momentum Scale. Must be a scalar.epsilon
- Ridge term. Must be a scalar.grad
- The gradient.options
- carries optional attribute valuespublic static ResourceApplyCenteredRmsProp.Options useLocking(Boolean useLocking)
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.Copyright © 2015–2022. All rights reserved.