Skip to content

Optuna

Optuna project

Optuna is a general Python framework for hyperparameter search with study, trial, sampling and pruning APIs.

Catalog updated ·

Overview

Optuna organizes optimization around studies and trials. Its documented Python API defines search spaces in objective functions and offers samplers, pruning of unpromising trials and visualization of study history. The project is designed primarily for machine-learning hyperparameter optimization.

Optuna is not a chemistry-specific laboratory controller. Chemical use requires a user-defined objective and safeguards for feasible inputs, costs and measurements. The repository documents Python package installation and an MIT license.

Key Features

  • Study and trial APIs for parameter search
  • Sampling, pruning and study-history visualization

Use Cases

  • Tune a molecular prediction model or explore a bounded experimental parameter space after defining an evaluation function.

How to Use

Install Optuna and run the official example. Specify the objective, search ranges and stopping rules; inspect trial records and validate any proposed condition before use.

Sources and verification

This entry describes capabilities documented by the linked project sources. Catalog edits are not evidence of a successful run, compatibility test or scientific performance evaluation.

Read the editorial and verification policy

Related resources

AlabOS

Open Source

AlabOS manages the workflows of an autonomous laboratory, coordinating samples, tasks and device actions through a database-backed orchestration service.

Open sourcePython

Materials Discovery · Lab Automation · Process Optimization

Atlas is a Bayesian optimization package for experimental science, providing Gaussian-process planners for self-driving-laboratory workflows.

Open sourcePython

Lab Automation · Process Optimization

Ax

Platforms

Ax is a Python platform for adaptive experiments, including Bayesian optimization powered by BoTorch.

Open sourcePython

Process Optimization

BayBE

Open Source

BayBE is a Python toolbox for Bayesian experimental design that recommends parameter configurations from defined search spaces, objectives and measurements, including chemistry-aware and multi-target workflows.

Open sourcePython

Lab Automation · Process Optimization

BoFire

Open Source

BoFire is a Python framework for experimental design and Bayesian optimization, supporting mixed variables, constraints, molecular representations and iterative candidate selection.

Open sourcePython

Lab Automation · Process Optimization

BoTorch

Open Source

BoTorch is a PyTorch library for constructing Bayesian optimization models, acquisition functions and candidate generation workflows.

Open sourcePython

Process Optimization

Related guides