What specific boosting implementations do you know or have worked with?

Main boosting implementations that are commonly used:

• Gradient Boosting Machines (GBM) — a classic gradient boosting, implemented, for example, in the scikit-learn library.
• XGBoost — a highly efficient and popular implementation of gradient boosting with optimizations for speed and memory.
• LightGBM — from Microsoft, using leaf-wise tree growth and optimizations for large data.
• CatBoost — from Yandex, works well with categorical features and has built-in handling of missing values.

Each of these implementations has its own features and advantages; the choice depends on the task and data.