Skip to content

OpenAI Gym environment for simulating a soccer match

Notifications You must be signed in to change notification settings

yc2454/gym-futbol

Repository files navigation

gym-futbol

• A reinforcement learning environment where agents are trained to play soccer in a primitive 2D simulator. AI players are trained to play with hard-coded opponents. Developed with API from OpenAI's Gym.

• Implemented different reinforcement algorithms for training, such as Deep Q learning, PPO(Proximal Policy Optimization) and Advantage Actor Critic (A2C)

Use the colab_notebook to try our environments and trained agents

About

OpenAI Gym environment for simulating a soccer match

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published