mxnet.optimizer.lamb

Lamb optimizer.

Classes

LAMB([learning_rate, beta1, beta2, epsilon, ...])

LAMB Optimizer.
class mxnet.optimizer.lamb.LAMB(learning_rate=0.001, beta1=0.9, beta2=0.999, epsilon=1e-06, lower_bound=None, upper_bound=None, bias_correction=True, aggregate_num=4, use_fused_step=True, **kwargs)[source]

Bases: Optimizer

LAMB Optimizer.

Referenced from ‘Large Batch Optimization for Deep Learning: Training BERT in 76 minutes’ (https://arxiv.org/pdf/1904.00962.pdf)

Parameters:

  • learning_rate (float, default 0.001) – The initial learning rate. If None, the optimization will use the learning rate from lr_scheduler. If not None, it will overwrite the learning rate in lr_scheduler. If None and lr_scheduler is also None, then it will be set to 0.01 by default.

  • beta1 (float, default 0.9) – Exponential decay rate for the first moment estimates.

  • beta2 (float, default 0.999) – Exponential decay rate for the second moment estimates.

  • epsilon (float, default 1e-6) – Small value to avoid division by 0.

  • lower_bound (float, default None) – Lower limit of norm of weight

  • upper_bound (float, default None) – Upper limit of norm of weight

  • bias_correction (bool, default True) – Whether or not to apply bias correction

  • aggregate_num (int, default 4) – Number of weights to be aggregated in a list. They are passed to the optimizer for a single optimization step. In default, all the weights are aggregated.

  • use_fused_step (bool, default True) – Whether or not to use fused kernels for optimizer. When use_fused_step=False, step is called, otherwise, fused_step is called.

create_state(index, weight)[source]

Creates auxiliary state for a given weight.

Some optimizers require additional states, e.g. as momentum, in addition to gradients in order to update weights. This function creates state for a given weight which will be used in update. This function is called only once for each weight.

Parameters:

  • index (int) – An unique index to identify the weight.

  • weight (NDArray) – The weight.

Returns:

state – The state associated with the weight.

Return type:

any obj

fused_step(indices, weights, grads, states)[source]

Perform a fused optimization step using gradients and states. Fused kernel is used for update.

Parameters:

  • indices (list of int) – List of unique indices of the parameters into the individual learning rates and weight decays. Learning rates and weight decay may be set via set_lr_mult() and set_wd_mult(), respectively.

  • weights (list of NDArray) – List of parameters to be updated.

  • grads (list of NDArray) – List of gradients of the objective with respect to this parameter.

  • states (List of any obj) – List of state returned by create_state().

step(indices, weights, grads, states)[source]

Perform a fused optimization step using gradients and states. Fused kernel is used for update.

Parameters:

  • indices (list of int) – List of unique indices of the parameters into the individual learning rates and weight decays. Learning rates and weight decay may be set via set_lr_mult() and set_wd_mult(), respectively.

  • weights (list of NDArray) – List of parameters to be updated.

  • grads (list of NDArray) – List of gradients of the objective with respect to this parameter.

  • states (List of any obj) – List of state returned by create_state().

update_multi_precision(indices, weights, grads, states)[source]

Override update_multi_precision.