ascentai / diy-gym

A framework for creating your own reinforcement learning environments using pybullet
MIT License
19 stars 3 forks source link

Add config to flatten observations and actions into spaces.Box #15

Closed thomascent closed 5 years ago

thomascent commented 5 years ago

Added configs to flatten observations and actions down into a Box space. Traversing dictionary spaces is kind of a pain and it seems like lots of RL implementations (looking at you baselines/ppo1) assume at least a Box observation space