Skip to content

Miniworld 2.0.1: Miniworld Becomes Mature

Latest
Compare
Choose a tag to compare
@jkterry1 jkterry1 released this 14 Feb 16:16
· 65 commits to master since this release
bd2ffb5

Release Notes

MiniWorld is a minimalistic 3D interior environment simulator for reinforcement learning & robotics research that allows environments to be easily edited like Minigrid meets DM Lab. It can simulate environments with rooms, doors, hallways, and various objects (e.g., office and home environments, mazes). Miniworld 2.0.0 is the first mature release within Farama. This version transitions from gym to gymnasium. Additionally, this release adds CI testing, code standardization pipeline, tests for environments, and documentation for each environment.

Furthermore, we have a website (https://miniworld.farama.org/) that has documentation that covers all the relevant details to start implementing a reinforcement learning agent. In future releases, we plan to add tutorials and more detailed documentation.

New Features and Improvements

  • Added Sign environment from the DREAM paper by @ezliu in #47
  • Updated to Gymnasium v0.26.2 by @BolunDai0216 in #72
  • Replaced random number generator from random.py with np_random from gymnasium by @hh2564 in #74
  • Added EzPickle inheritance to the environments which enables pickling and unpickling objects via their constructor arguments by @BolunDai0216 in #76

Bug Fixes and Documentation Updates

  • Made offscreen gym-miniworld work without updating NVIDIA drivers by @ptigas in #40
  • Added docstrings for the environments and updated manual_control.py by @BolunDai0216 in #73
  • Updated README and manual_control.py by @BolunDai0216 in #77
  • Added docs website by @mgoulao in #78
  • Updated docstrings for all of the environments by @BolunDai0216 in #81