MVA-2021 / reinforcement_learning / hw1_dynamic_programming / vi_cvg.png