Meta-World+: An Improved, Standardized, RL Benchmark
Alberta Machine Intelligence Institute Toronto Metropolitan University · University of Surrey · Hamburg University of Technology · Stanford University · Columbia University · University of Southern California · Google · University of Maryland College Park (SSO) · Toronto Metropolitan University · Google DeepMind / U. de Montreal / Mila
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
Meta-World is widely used for evaluating multi-task and meta-reinforcement learning agents, which are challenged to master diverse skills simultaneously. Since its introduction however, there have been numerous undocumented changes which inhibit a fair comparison of algorithms. This work strives to disambiguate these results from the literature, while also leveraging the past versions of Meta-World to provide insights into multi-task and meta-reinforcement learning benchmark design. Through this process we release an open-source version of Meta-World that has full reproducibility of past results, is more technically ergonomic, and gives users more control over the tasks that are included in a task set.