[~Re] Exploration in Model-based Reinforcement Learning by Empirically Estimating Learning Progress