Optimizer

class qctrl.dynamic.types.closed_loop_optimization_step.Optimizer(*, cross_entropy_initializer=None, gaussian_process_initializer=None, neural_network_initializer=None, state=None)

The optimizer to use in the optimization. Exactly one field must be non-null. For the first step of the optimizer, this must be an initializer, such as a CrossEntropyInitializer. For subsequent steps, you need to provide the state returned in the Result of the previous step, which already contains the optimizer configuration and previous test points.

Variables

• cross_entropy_initializer (qctrl.dynamic.types.closed_loop_optimization_step.CrossEntropyInitializer) – Initializer for the cross-entropy optimizer. Pass this field to use this algorithm for the closed-loop optimization.

• gaussian_process_initializer (qctrl.dynamic.types.closed_loop_optimization_step.GaussianProcessInitializer) – Initializer for the optimizer based on Gaussian processes. Pass this field to use this algorithm for the closed-loop optimization.

• neural_network_initializer (qctrl.dynamic.types.closed_loop_optimization_step.NeuralNetworkInitializer) – Initializer for the optimizer based on neural networks. Pass this field to use this algorithm for the closed-loop optimization.

• state (str) – The state of the optimizer at the current step, as returned in the Result of the previous step. Must be updated in following steps after the optimizer is initialized.