Hyperparameter Tuner Skill for GridSearchCV Comparisons
Define a small search grid and fixed cross-validation so model settings can be compared reproducibly.
中文Install in Claude Code
Use Python, scikit-learn, and a synthetic classification fixture. Reserve a test split for the final evaluation.
- Get the complete `plugins/ai-ml/hyperparameter-tuner/skills/tuning-hyperparameters` directory from the pinned GitHub commit.
- Copy it into the target project's `.claude/skills/`. Check `.claude/skills/tuning-hyperparameters/SKILL.md`, start Claude Code at the project root, and call `/tuning-hyperparameters`.
Example task
Compare random forest max_depth values 2, 4, and 6 on synthetic catalog classes. Fix the model and stratified fold seeds. Select by macro F1, then evaluate once on holdout.What to check
The search table shows all three scores, and the holdout report identifies the selected configuration.
Original SKILL.md on GitHub · Updated
Frequently asked question
Can I revise the grid based on the final test score?
Keep the test set for the end. Use validation within training to decide whether to widen the grid.