Gym wrapper for DeepMind Lab environments

Gym implementation of connector to Deepmind lab

Getting Started

Install the newest package by running:

pip install gym_deepmindlab

In your project:

import gym
import gym_deepmindlab

env = gym.make('DeepmindLabSeekavoidArena01-v0')

# Use the environment
observation = env.reset()

Suppoted Environments

This is the list of supported environments:

• DeepmindLabLtChasm-v0
• DeepmindLabLtHallwaySlope-v0
• DeepmindLabLtHorseshoeColor-v0
• DeepmindLabLtSpaceBounceHard-v0
• DeepmindLabNavMazeRandomGoal01-v0
• DeepmindLabNavMazeRandomGoal02-v0
• DeepmindLabNavMazeRandomGoal03-v0
• DeepmindLabNavMazeStatic01-v0
• DeepmindLabNavMazeStatic02-v0
• DeepmindLabSeekavoidArena01-v0
• DeepmindLabStairwayToMelon-v0

Arguments

You can specify the screen size when creating a new environment:

import gym
import gym_deepmindlab

env = gym.make('DeepmindLabSeekavoidArena01-v0', width = 224, height = 224)

# Use the environment
observation = env.reset()

Thanks

Thanks to https://github.com/deepmind/lab for such a great work.