| Property | Value |
|---|---|
| Author | ThomasSimonini |
| Training Steps | 5.1M |
| Final ELO Score | 1766.452 |
| Framework | Unity ML-Agents |
| Repository | Hugging Face |
What is ML-Agents-SnowballFight-1vs1?
ML-Agents-SnowballFight-1vs1 is a sophisticated multi-agent reinforcement learning environment built using Unity ML-Agents Toolkit. It simulates a competitive 1v1 snowball fight where agents must learn both offensive and defensive strategies to succeed. The environment demonstrates practical applications of multi-agent reinforcement learning in competitive scenarios.
Implementation Details
The model utilizes PPO (Proximal Policy Optimization) with sophisticated observation and action spaces. The neural network architecture features 2 hidden layers with 512 units each, trained with a batch size of 2048 and buffer size of 20480. The implementation includes self-play mechanisms for continuous improvement, with team changes every 200,000 steps and a window of 10 for opponent selection.
- Ray-cast based perception system with 33 rays for environment sensing
- Discrete action space with branched actions for movement and shooting
- Self-play training with ELO rating system starting at 1200
- Reward shaping based on hit success and time penalties
Core Capabilities
- Complex spatial reasoning through ray-cast observations
- Strategic decision making for shooting and movement
- Health and cooldown management
- Advanced opponent tracking and avoidance
Frequently Asked Questions
Q: What makes this model unique?
This model uniquely combines competitive gameplay with sophisticated multi-agent learning, featuring a complex observation space and carefully crafted reward system that promotes both offensive and defensive strategies.
Q: What are the recommended use cases?
The model is ideal for studying multi-agent reinforcement learning, competitive AI behavior, and can serve as a foundation for developing more complex game AI systems or studying adversarial behaviors in contained environments.