Module PINNacle.src.optimizer.multiadam

Functions

sadam(params: List[torch.Tensor], grads: List[List[torch.Tensor]], exp_avgs: List[List[torch.Tensor]], exp_avg_sqs: List[List[torch.Tensor]], max_exp_avg_sqs: List[List[torch.Tensor]], agg_exp_avg: List[torch.Tensor], agg_exp_avg_sqs: List[torch.Tensor], state_steps: List[int], *, amsgrad: bool, beta1: float, beta2: float, lr: float, weight_decay: float, eps: float, maximize: bool, group_weights: torch.Tensor, agg_momentum: bool, agg_beta1: float, agg_beta2: float) : Functional API that performs MultiAdam algorithm computation.

See :class:`~torch.optim.Adam` for details.

Classes

MultiAdam(params, lr=0.001, betas=(0.99, 0.99), eps=1e-08, weight_decay=0, amsgrad=False, maximize=False, loss_group_idx=None, group_weights=None, agg_momentum=False, agg_betas=None, *, param_scheduler=None) : Base class for all optimizers.

.. warning::
    Parameters need to be specified as collections that have a deterministic
    ordering that is consistent between runs. Examples of objects that don't
    satisfy those properties are sets and iterators over values of dictionaries.

Args:
    params (iterable): an iterable of :class:`torch.Tensor` s or
        :class:`dict` s. Specifies what Tensors should be optimized.
    defaults: (dict): a dict containing default values of optimization
        options (used when a parameter group doesn't specify them).

### Ancestors (in MRO)

* torch.optim.optimizer.Optimizer

### Methods

`init_states(self)`
:

`step(self, closure)`
:   Performs a single optimization step.

    Args:
        closure (callable, optional): A closure that reevaluates the model
            and returns the loss.

ParamScheduler(epochs=20000, lr_scheduler=None, betas_scheduler=None, group_weights_scheduler=None, default_lr=0.001, default_betas=(0.99, 0.99), default_group_weights=(0.5, 0.5)) :

### Methods

`betas(self)`
:

`group_weights(self)`
:

`lr(self)`
:

`step(self, losses, grouped_losses)`
:

Lr Adaptor Ntk