Train and deploy CatBoost 1.2.10 gradient boosting models for classification, regression, ranking, and multi-output tasks. Use when the user needs gradient boosting on tabular data with categorical features, GPU acceleration, text features, embedding features, monotone constraints, SHAP values, or model export to ONNX/CoreML/PMML/C++. Supports scikit-learn API and CatBoost-specific API (Pool, train, cv). Handles missing values natively. Apache 2.0 license.