Skip to content

v1.0.0

Latest
Compare
Choose a tag to compare
@chrehall68 chrehall68 released this 07 Mar 16:08
00afd92

This version contains working code to train a PPO agent to play the snake game. A pretrained model is also provided, and it can be run with ai_player.py

Features

  • PPO algorithm code, written by hand to follow the PPO paper
  • Snake env, a gymnasium environment that simulates the snake game
  • Generic Actor-Critic policy networks, and a specialized network specifically for the snake game