Skip to content

An example project using vel to train reinforcement learning agents on existing community gym environments. A work in progress repository.

License

Notifications You must be signed in to change notification settings

MillionIntegrals/vel-miniworld

Folders and files

NameName
Last commit message
Last commit date

Latest commit

1078fbe · Nov 25, 2018

History

5 Commits
Nov 25, 2018
Nov 25, 2018
Nov 25, 2018
Nov 24, 2018
Nov 24, 2018
Nov 24, 2018
Nov 23, 2018
Nov 25, 2018
Nov 25, 2018
Nov 25, 2018

Repository files navigation

vel-miniworld

An example project using vel to train reinforcement learning agents on existing community gym environments. A work-in-progress repository.

Supported environments:

Examples confirmed to be working:

  • examples-configs/ppo/ppo_minigrid_empty_8x8.yaml
  • examples-configs/ppo/ppo_minigrid_doorkey_6x6.yaml (doesn't converge every time)
  • examples-configs/ppo/ppo_miniworld_hallway.yaml

How to run?

git clone git@github.com:MillionIntegrals/vel-miniworld.git
cd vel-miniworld

# Optionally, if you don't want to store metrics in the db and visualize in VisDom
mv .velproject.dummy.yaml .velproject.yaml

pipenv install
pipenv shell
vel examples-configs/ppo/ppo_minigrid_empty_8x8.yaml train
vel examples-configs/ppo/ppo_minigrid_empty_8x8.yaml record

# Optionally, play a video of agent solving a rather simple environment
mplayer output/videos/ppo_minigrid_empty_8x8/0/ppo_minigrid_empty_8x8_vid_0010.avi

Additional notes

For the textures to load properly for the 3D rendered miniworld environment, it needs to be installed from a git repository, by running pip install -e . in the top-level directory of the checkout.

Let me know if you have any other problems running the environments.

Some animations

Solving simple small gridworld environment:

Solving slightly more complex gridworld environment with sparse rewards:

Solving small 3D rendered world:

About

An example project using vel to train reinforcement learning agents on existing community gym environments. A work in progress repository.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages