google-research/football
A reinforcement learning environment simulating football games for training autonomous game-playing agents.

Velocity · 7d
+1.4
★ / day
Trend
→steady
star history
Google Research Football provides an open-source RL benchmark environment based on the Gameplay Football simulator. Researchers can train agents using popular RL algorithms to play the game, with support for custom scenarios, multi-agent gameplay, and diverse observation formats. The repository includes pre-built environments, training examples, and is widely used for RL research and competitions including a Kaggle competition.