OptunaHub

Registry of the third-party Optuna packages.

The optunahub library is required to use the packages in this registry.
You can publish your package via optunahub-registry. See the OptunaHub tutorial.

TPE Sampler
Sampler using TPE (Tree-structured Parzen Estimator) algorithm.
Sampler Built-In

Timeline Plot
Plot the timeline of a study.
Visualization Built-In

Terminator Improvement Plot
Plot the potentials for future objective improvement.
Visualization Built-In

Slice Plot
Plot the parameter relationship as slice plot in a study.
Visualization Built-In

Simulated Annealing Sampler
Sampler based on simulated annealing algorithm.
Sampler Simulated Annealing

Simple Sampler
An easy sampler base class to implement custom samplers.
Sampler Development

Sampler using Whale Optimization Algorithm
Swarm Algorithm Inspired by Pod of Whale
Sampler

Rank Plot
Plot parameter relations as scatter plots with colors indicating ranks of target value.
Visualization Built-In

Random Search
Sampler using random sampling.
Sampler Built-In

QMC Search
Sampler using Quasi Monte Carlo sampling.
Sampler Built-In

PyCMA Sampler
A CMA-ES Sampler using cma library as the backend.
Sampler Built-In

Plot Hypervolume History with Reference Point
Plot hypervolume history with the reference point information.
Visualization Hypervolume Multi-Objective Optimization

PLMBO (Preference Learning Multi-Objective Bayesian Optimization)
Interatctive multi-objective Bayesian optimization based on user preference
Sampler Interactive Optimization Bayesian Optimization Multi-Objective Optimization Preference Learning Active Learning

PFNs4BO sampler
In-context learning for Bayesian optimization. This sampler uses Prior-data Fitted Networks (PFNs) as a surrogate model for Bayesian optimization.
Sampler Bayesian Optimization Pre-Trained Model

Percentile Pruner
Pruner to keep the specified percentile of the trials.
Pruner Built-In

Partial Fixed Sampler
Sampler with partially fixed parameters.
Sampler Built-In

Pareto-front Plot
Plot the Pareto front of a study.
Visualization Built-In

Parallel Coordinate Plot
Plot the high-dimensional parameter relationships in a study.
Visualization Built-In

Optimization History Plot
Plot optimization history of all trials in a study.
Visualization Built-In

NSGAIII Search
Sampler using NSGAIII algotithm.
Sampler Built-In

NSGAII Search
Sampler using NSGAII algotithm.
Sampler Built-In

Median Pruner
Pruner using the median stopping rule.
Pruner Built-In

Intermediate Values Plot
Plot intermediate values of all trials in a study.
Visualization Built-In

Implicit Natural Gradient Sampler
A sampler based on Implicit Natural Gradient.
Sampler

Hypervolume History Plot
Plot hypervolume history of all trials in a study.
Visualization Built-In

Hyperparameter Importances Plot
Plot hyperparameter importances.
Visualization Built-In

HEBO (Heteroscedastic and Evolutionary Bayesian Optimisation)
HEBO addresses the problem of noisy and heterogeneous objective functions by using a heteroscedastic Gaussian process and an evolutionary algorithm.
Sampler Bayesian Optimization Heteroscedastic Gaussian Process Evolutionary Algorithm

Grid Search
Sampler using grid search.
Sampler Built-In

Gaussian Process-Based Sampler
Sampler using Gaussian process-based Bayesian optimization.
Sampler Built-In

Evolutionary LLM Merge Sampler
A sampler for evolutionary LLM merge.
Sampler LLM

Empirical Distribution Function Plot
Plot the objective value EDF (empirical distribution function) of a study.
Visualization Built-In

Demo Sampler
Demo Sampler of OptunaHub
Sampler

Contour Plot
Plot the parameter relationship as contour plot in a study.
Visualization Built-In

CMA-ES Sampler
A sampler using cmaes as the backend.
Sampler Built-In

Brute Force Search
Sampler using brute force.
Sampler Built-In

BoTorch Sampler
A Sampler using botorch library as the backend.
Sampler Built-In