Package list

Packages for blackbox optimization

Python packages for optimization using black-box search, gradient-free optimization and metaheuristics such as evolutionary algorithms.

Package	Comment	Downloads/day	Recent releases ⓘ	GitHub Stars
optuna	A hyperparameter and black-box optimization framework using adaptive sampling strategies such as TPE and CMA-ES, with built-in pruning and a web dashboard.	354K	6	·
pymoo	A comprehensive framework for multi-objective optimization, featuring NSGA-II, NSGA-III, and many other evolutionary algorithms with visualization support.	17K	1	·
deap	A flexible evolutionary computation framework that supports genetic algorithms, genetic programming, evolution strategies, and particle swarm optimization.	11K	1	·
nevergrad	Meta's gradient-free optimization library with a large collection of algorithms (CMA-ES, DE, PSO, etc.) designed for noisy and high-dimensional black-box problems.	11K	0	·
hyperopt	Distributed hyperparameter optimization using random search and Tree of Parzen Estimators (TPE), with MongoDB back-end support for parallel trials.	99K	0	·

keras-tunerA Hyperparameter Tuning Library for Keras
scikit-optimizeSequential model-based optimization toolbox.
optaxA gradient processing and optimization library in JAX.
cvxoptConvex optimization package
flamlA fast library for automated machine learning and tuning
cvxpyA domain-specific language for modeling convex optimization problems in Python.
peftParameter-Efficient Fine-Tuning (PEFT)
tensorflow-model-optimizationA suite of tools that users, both novice and advanced can use to optimize machine learning models for deployment and execution.
pyomoThe Pyomo optimization modeling framework
skopsA set of tools, related to machine learning in production.

keras-tunerA Hyperparameter Tuning Library for Keras
scikit-optimizeSequential model-based optimization toolbox.
optaxA gradient processing and optimization library in JAX.
cvxoptConvex optimization package
flamlA fast library for automated machine learning and tuning
cvxpyA domain-specific language for modeling convex optimization problems in Python.
peftParameter-Efficient Fine-Tuning (PEFT)
tensorflow-model-optimizationA suite of tools that users, both novice and advanced can use to optimize machine learning models for deployment and execution.
pyomoThe Pyomo optimization modeling framework
skopsA set of tools, related to machine learning in production.

Package	Comment	Downloads/day	Recent releases ⓘ	GitHub Stars
optuna	A hyperparameter and black-box optimization framework using adaptive sampling strategies such as TPE and CMA-ES, with built-in pruning and a web dashboard.	354K	6	11,707·
pymoo	A comprehensive framework for multi-objective optimization, featuring NSGA-II, NSGA-III, and many other evolutionary algorithms with visualization support.	17K	1	—·
deap	A flexible evolutionary computation framework that supports genetic algorithms, genetic programming, evolution strategies, and particle swarm optimization.	11K	1	—·
nevergrad	Meta's gradient-free optimization library with a large collection of algorithms (CMA-ES, DE, PSO, etc.) designed for noisy and high-dimensional black-box problems.	11K	0	4,160·
hyperopt	Distributed hyperparameter optimization using random search and Tree of Parzen Estimators (TPE), with MongoDB back-end support for parallel trials.	99K	0	—·