hpo

What problem does it solve?

This Skill automates the process of finding the optimal hyperparameters for machine learning models, saving significant manual effort and computational resources.

Core Features & Use Cases

• Bayesian Optimization: Leverages Optuna's TPE sampler for efficient hyperparameter search.
• Fold-Level Pruning: Implements MedianPruner to cut off unpromising trials early, saving compute.
• Model Support: Provides templates for LightGBM, XGBoost, and CatBoost.
• Persistence: Saves results to SQLite, allowing interrupted searches to be resumed.
• Use Case: After establishing a competitive baseline model, use this Skill to fine-tune its hyperparameters to achieve the best possible performance on your dataset.

Quick Start

Run the hyperparameter optimization script for LightGBM using your data.