tf.raw_ops.ApplyFtrl

Update '*var' according to the Ftrl-proximal scheme.

tf.raw_ops.ApplyFtrl(
    var,
    accum,
    linear,
    grad,
    lr,
    l1,
    l2,
    lr_power,
    use_locking=False,
    multiply_linear_by_lr=False,
    name=None
)

accum_new = accum + grad * grad linear += grad - (accum_new^(-lr_power) - accum^(-lr_power)) / lr * var quadratic = 1.0 / (accum_new^(lr_power) * lr) + 2 * l2 var = (sign(linear) * l1 - linear) / quadratic if |linear| > l1 else 0.0 accum = accum_new

varA mutable Tensor. Must be one of the following types: float32, float64, int32, uint8, int16, int8, complex64, int64, qint8, quint8, qint32, bfloat16, qint16, quint16, uint16, complex128, half, uint32, uint64. Should be from a Variable().
accumA mutable Tensor. Must have the same type as var. Should be from a Variable().
linearA mutable Tensor. Must have the same type as var. Should be from a Variable().
gradA Tensor. Must have the same type as var. The gradient.
lrA Tensor. Must have the same type as var. Scaling factor. Must be a scalar.
l1A Tensor. Must have the same type as var. L1 regularization. Must be a scalar.
l2A Tensor. Must have the same type as var. L2 regularization. Must be a scalar.
lr_powerA Tensor. Must have the same type as var. Scaling factor. Must be a scalar.
use_lockingAn optional bool. Defaults to False. If True, updating of the var and accum tensors will be protected by a lock; otherwise the behavior is undefined, but may exhibit less contention.
multiply_linear_by_lrAn optional bool. Defaults to False.
nameA name for the operation (optional).

A mutable Tensor. Has the same type as var.