GAROM#

class GAROM(problem, generator, discriminator, loss=None, optimizer_generator=None, optimizer_discriminator=None, scheduler_generator=None, scheduler_discriminator=None, gamma=0.3, lambda_k=0.001, regularizer=False)[source]#

Bases: MultiSolverInterface

GAROM solver class. This class implements Generative Adversarial Reduced Order Model solver, using user specified models to solve a specific order reduction problem.

See also

Original reference: Coscia, D., Demo, N., & Rozza, G. (2023). Generative Adversarial Reduced Order Modelling. DOI: arXiv preprint arXiv:2305.15881..

Initialization of the GAROM class.

Parameters:

  • problem (AbstractProblem) – The formulation of the problem.

  • generator (torch.nn.Module) – The generator model.

  • discriminator (torch.nn.Module) – The discriminator model.

  • loss (torch.nn.Module) – The loss function to be minimized. If None, PowerLoss with p=1 is used. Default is None.

  • optimizer_generator (Optimizer) – The optimizer for the generator. If None, the torch.optim.Adam optimizer is used. Default is None.

  • optimizer_discriminator (Optimizer) – The optimizer for the discriminator. If None, the torch.optim.Adam optimizer is used. Default is None.

  • scheduler_generator (Scheduler) – The learning rate scheduler for the generator. If None, the torch.optim.lr_scheduler.ConstantLR scheduler is used. Default is None.

  • scheduler_discriminator (Scheduler) – The learning rate scheduler for the discriminator. If None, the torch.optim.lr_scheduler.ConstantLR scheduler is used. Default is None.

  • gamma (float) – Ratio of expected loss for generator and discriminator. Default is 0.3.

  • lambda_k (float) – Learning rate for control theory optimization. Default is 0.001.

  • regularizer (bool) – If True, uses a regularization term in the GAROM loss. Default is False.

accepted_conditions_types#

alias of InputTargetCondition

forward(x, mc_steps=20, variance=False)[source]#

Forward pass implementation.

Parameters:

  • x (torch.Tensor) – The input tensor.

  • mc_steps (int) – Number of Montecarlo samples to approximate the expected value. Default is 20.

  • variance (bool) – If True, the method returns also the variance of the solution. Default is False.

Returns:

The expected value of the generator distribution. If variance=True, the method returns also the variance.

Return type:

torch.Tensor | tuple[torch.Tensor, torch.Tensor]

sample(x)[source]#

Sample from the generator distribution.

Parameters:

x (torch.Tensor) – The input tensor.

Returns:

The generated sample.

Return type:

torch.Tensor

on_train_batch_end(outputs, batch, batch_idx)[source]#

This method is called at the end of each training batch and overrides the PyTorch Lightning implementation to log checkpoints.

Parameters:

  • outputs (torch.Tensor) – The model’s output for the current batch.

  • batch (list[tuple[str, dict]]) – A batch of data. Each element is a tuple containing a condition name and a dictionary of points.

  • batch_idx (int) – The index of the current batch.

optimization_cycle(batch)[source]#

The optimization cycle for the GAROM solver.

Parameters:

batch (list[tuple[str, dict]]) – A batch of data. Each element is a tuple containing a condition name and a dictionary of points.

Returns:

The losses computed for all conditions in the batch, casted to a subclass of torch.Tensor. It should return a dict containing the condition name and the associated scalar loss.

Return type:

dict

validation_step(batch)[source]#

The validation step for the PINN solver.

Parameters:

batch (list[tuple[str, dict]]) – A batch of data. Each element is a tuple containing a condition name and a dictionary of points.

Returns:

The loss of the validation step.

Return type:

torch.Tensor

test_step(batch)[source]#

The test step for the PINN solver.

Parameters:

batch (list[tuple[str, dict]]) – A batch of data. Each element is a tuple containing a condition name and a dictionary of points.

Returns:

The loss of the test step.

Return type:

torch.Tensor

property generator#

The generator model.

Returns:

The generator model.

Return type:

torch.nn.Module

property discriminator#

The discriminator model.

Returns:

The discriminator model.

Return type:

torch.nn.Module

property optimizer_generator#

The optimizer for the generator.

Returns:

The optimizer for the generator.

Return type:

Optimizer

property optimizer_discriminator#

The optimizer for the discriminator.

Returns:

The optimizer for the discriminator.

Return type:

Optimizer

property scheduler_generator#

The scheduler for the generator.

Returns:

The scheduler for the generator.

Return type:

Scheduler

property scheduler_discriminator#

The scheduler for the discriminator.

Returns:

The scheduler for the discriminator.

Return type:

Scheduler