This repository provides the gym-soccer environment, a multi-agent domain with continuous state and action spaces designed for reinforcement learning research. It offers three distinct tasks: scoring a goal against an empty net with sparse rewards, scoring with more informative dense rewards, and scoring against a hand-coded goalkeeper.

How It Works

The environment simulates a soccer game with continuous state and action spaces, allowing agents to learn complex behaviors like ball control and shooting. The tasks vary in reward structure, from sparse goal-scoring rewards to denser rewards for ball proximity and movement towards the goal, catering to different learning challenges.

Quick Start & Requirements

• Install via pip install -e . after navigating to the gym-soccer directory.
• Requires Python.

Highlighted Details

• Continuous state and action spaces.
• Three tasks: Soccer (sparse reward), SoccerEmptyGoal (dense reward), SoccerAgainstKeeper (with a hand-coded goalkeeper).

Maintenance & Community

The project is archived and no updates are expected.

Licensing & Compatibility

The license is not specified in the README.

Limitations & Caveats

The project is archived and will not receive further updates. The specific license is not stated, which may impact commercial use.