best_model_params.json

{
    "n_estimators": 300,
    "min_samples_split": 2,
    "min_samples_leaf": 1,
    "max_features": "log2",
    "max_depth": null
}