SearchAlgorithms

The following search algorithms are available in Olive.

Each algorithm is followed by a description of the algorithm and a list of the its configuration options.

ExhaustiveSearchAlgorithm

Name: "exhaustive"

Exhaustive Search Algorithm. Does a grid search over the search space.

Name: "random"

Random Searcher. Samples random points from the search space with or without replacement.

num_samples

Number of samples to suggest.

type: int

default_value: 1

seed

Seed for the rng.

type: int

default_value: 1

with_replacement

Sample with replacement.

type: bool

default_value: False

Name: "tpe"

Sample using TPE (Tree-structured Parzen Estimator) algorithm. Uses optuna TPESampler underneath. Refer to https://optuna.readthedocs.io/en/stable/reference/samplers/generated/optuna.samplers.TPESampler.html for more details about the sampler.

num_samples

Number of samples to suggest.

type: int

default_value: 1

seed

Seed for the rng.

type: int

default_value: 1

multivariate

Use multivariate TPE when suggesting parameters.

type: bool

default_value: True

group

If this and multivariate are True, the multivariate TPE with the group decomposed search space is used when suggesting parameters. Refer to ‘group’ at https://optuna.readthedocs.io/en/stable/reference/samplers/generated/optuna.samplers.TPESampler.html for more information.

type: bool

default_value: True