They are the original codes (2020) used for the paper
POMO: Policy Optimization with Multiple Optima for Reinforcement Learning
accepted at NeurIPS 2020
http://arxiv.org/abs/2010.16011 https://proceedings.neurips.cc/paper/2020/hash/f231f2107df69eab0a3862d50018a9b2-Abstract.html
They are based on ipynb files and easier to play with interactively.
They are the updated codes, newly structured in 2021.
They are based on py files, so that they can be run on servers more easily.