snake and DRL4SnakeGame
These are ecosystem siblings—one implements a general Snake game environment while the other applies a specific deep reinforcement learning algorithm (DRL) to solve it, where DRL4SnakeGame likely uses or parallels the game mechanics defined in chynl/snake.
About snake
chynl/snake
Artificial intelligence for the Snake game.
Implements three distinct solver algorithms—Hamilton cycle pathfinding for near-optimal play, greedy heuristic search, and experimental deep Q-learning—evaluated across 1000-episode trials measuring final snake length and step efficiency. Built in Python with Tkinter visualization and includes comprehensive unit tests for algorithm validation.
About DRL4SnakeGame
ZYunfeii/DRL4SnakeGame
Using deep reinforcement learning to play Snake game(贪吃蛇).
Implements PPO (Proximal Policy Optimization) for discrete action spaces with a custom PyTorch neural network architecture, achieving convergence in approximately 30 minutes of training. The project includes a pygame-based Snake environment simulator, reward visualization via matplotlib/seaborn, and modular separation between the RL agent, network architecture, and game environment for easy extension or adaptation to other discrete control problems.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work