@Operator(group="train") public final class ResourceApplyAdadelta extends PrimitiveOp
accum = rho() * accum + (1 - rho()) * grad.square(); update = (update_accum + epsilon).sqrt() * (accum + epsilon()).rsqrt() * grad; update_accum = rho() * update_accum + (1 - rho()) * update.square(); var -= update;
Modifier and Type | Class and Description |
---|---|
static class |
ResourceApplyAdadelta.Options
Optional attributes for
ResourceApplyAdadelta |
operation
Modifier and Type | Method and Description |
---|---|
static <T> ResourceApplyAdadelta |
create(Scope scope,
Operand<?> var,
Operand<?> accum,
Operand<?> accumUpdate,
Operand<T> lr,
Operand<T> rho,
Operand<T> epsilon,
Operand<T> grad,
ResourceApplyAdadelta.Options... options)
Factory method to create a class wrapping a new ResourceApplyAdadelta operation.
|
static ResourceApplyAdadelta.Options |
useLocking(Boolean useLocking) |
equals, hashCode, op, toString
public static <T> ResourceApplyAdadelta create(Scope scope, Operand<?> var, Operand<?> accum, Operand<?> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdadelta.Options... options)
scope
- current scopevar
- Should be from a Variable().accum
- Should be from a Variable().accumUpdate
- Should be from a Variable().lr
- Scaling factor. Must be a scalar.rho
- Decay factor. Must be a scalar.epsilon
- Constant factor. Must be a scalar.grad
- The gradient.options
- carries optional attributes valuespublic static ResourceApplyAdadelta.Options useLocking(Boolean useLocking)
useLocking
- If True, updating of the var, accum and update_accum tensors will be protected by
a lock; otherwise the behavior is undefined, but may exhibit less contention.Copyright © 2022. All rights reserved.