skills.rest
hung-phan
hung-phanCOMMUNITY
Explore all skills by this author →
GitHub

gradient-free-optimization

Optimize non-differentiable machine learning objectives using CMA-ES, PSO, and Bayesian Optimization.

1|Updated Jun 4, 2026
Search Tags:#optimization#machine learning#hyperparameter tuning#black-box#gradient-free
One-click install
npx skills add https://github.com/hung-phan/ml-skills --skill gradient-free-optimization
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: gradient-free-optimization
Source: https://github.com/hung-phan/ml-skills/tree/main/skills/ml-review/references/ml-training/gradient-free-optimization
Command: npx skills add https://github.com/hung-phan/ml-skills --skill gradient-free-optimization

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires cma, pyswarms, optuna, scikit-optimize, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge of optimizing complex machine learning problems where traditional gradient-based methods are not applicable or effective.

Core Features & Use Cases

  • Gradient-Free Optimization: Offers a suite of algorithms for non-differentiable or noisy objectives.
  • Use Cases: Ideal for hyperparameter optimization, neural architecture search, prompt optimization, reinforcement learning, and combinatorial problems.
  • Example: Optimize the hyperparameters of a machine learning model without relying on gradient information.

Quick Start

Use the gradient-free-optimization skill to optimize the learning rate of your model.

Frequently Asked Questions about gradient-free-optimization

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I optimize machine learning hyperparameters for non-differentiable objectives?▼

Gradient-free optimization handles non-differentiable or noisy machine learning objectives by applying algorithms like CMA-ES, PSO, and Bayesian Optimization to effectively tune hyperparameters without relying on gradient information.

What is gradient-free optimization used for in machine learning?▼

Gradient-free optimization is used for hyperparameter optimization, neural architecture search, prompt optimization, reinforcement learning, and combinatorial problems where traditional gradient-based methods are ineffective or inapplicable.

Can I use CMA-ES and PSO for neural architecture search?▼

Yes, you can use CMA-ES and PSO for neural architecture search. These gradient-free algorithms are provided to optimize complex model structures and combinatorial problems where gradients are unavailable or noisy.

Does Bayesian optimization work for noisy reinforcement learning environments?▼

Yes, Bayesian optimization works for noisy reinforcement learning environments. It is included as a gradient-free method to effectively optimize objectives where gradient-based approaches fail due to noise or non-differentiability.

When should I avoid gradient-based methods and use gradient-free optimization?▼

You should use gradient-free optimization instead of gradient-based methods when dealing with non-differentiable objectives, noisy environments, or complex combinatorial problems like prompt optimization and neural architecture search.

Related Skills

Explore More Software Engineering Skills→
yu13130122297yu13130122297

scout

Identify uncertainties in research framing, datasets, metrics, and baselines.

Community
Intermediate
Axel-DaMageAxel-DaMage
2

fine-tuning-expert

Configure LoRA/QLoRA adapters and hyperparameters for fine-tuning large language models.

Community
Advanced
nahisahonahisaho
69

ai-ml-engineer

Develop end-to-end ML models with training, evaluation, deployment, and MLOps.

Community
Advanced
pengqianhanpengqianhan

nevergrad-skill

Optimize black-box parameters using Nevergrad evolutionary algorithms.

Community
Advanced