ocatariwrappers

This repository includes different wrappers to be used with OCAtari to generate different object-centric masked input representations.

https://github.com/vanillawhey/ocatariwrappers

Science Score: 36.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
  • codemeta.json file
    Found codemeta.json file
  • .zenodo.json file
    Found .zenodo.json file
  • DOI references
  • Academic publication links
    Links to: arxiv.org
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (5.8%) to scientific vocabulary

Keywords

gymnasium masking ocatari
Last synced: 10 months ago · JSON representation

Repository

This repository includes different wrappers to be used with OCAtari to generate different object-centric masked input representations.

Basic Info
  • Host: GitHub
  • Owner: VanillaWhey
  • License: mit
  • Language: Python
  • Default Branch: main
  • Homepage:
  • Size: 255 MB
Statistics
  • Stars: 1
  • Watchers: 2
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Topics
gymnasium masking ocatari
Created over 1 year ago · Last pushed 11 months ago
Metadata Files
Readme License Citation

README.md

OCCAM-Wrappers

This is the official repository for the paper Deep Reinforcement Learning via Object-Centric Attention.

This repository includes different wrappers to be used with OCAtari to generate different object-centric masked input representations.

Install

pip install "gymnasium[atari, accept-rom-license]" pip install -r requirements.txt pip install .

Usage

```python from ocatari_wrappers import BinaryMaskWrapper from ocatari import OCAtari

env = OCAtari("ALE/Frostbite")

env = BinaryMaskWrapper(env)

obs, info = env.reset()

done = False while not done: action = env.action_space.sample() obs, reward, terminated, truncated, info = env.step(action) done = terminated or truncated ```

Test Setup

First test if the Backend is set up correctly

python scripts/run.py -g Pong -hu

Now we test if the wrappers are also set up

python scripts/print_state.py

If everything works as intended you should now have an svg showing you the binary mask in the game of Freeway after 100 steps.

Citing

Please cite as stated.

Owner

  • Name: Cedric Derstroff
  • Login: VanillaWhey
  • Kind: user
  • Location: Germany
  • Company: Technische Universität Darmstadt

PhD Student @ TU Darmstadt

GitHub Events

Total
  • Watch event: 2
  • Push event: 11
Last Year
  • Watch event: 2
  • Push event: 11

Dependencies

requirements.txt pypi
  • gymnasium ==0.29.1
  • hackatari ==0.0.6
  • matplotlib ==3.5.3
  • numpy ==1.24.4
  • ocatari ==2.1.0
  • opencv-python ==4.7.0.72
  • pygame ==2.1.0
  • rliable ==1.2.0
  • stable_baselines3 ==2.5.0
  • torch ==2.6.0
setup.py pypi