This repository is to develop RL algorithms for sokoban env
A-start algorithm is for benchmark and test if the result is possible.
- get the state of the all the boxes and worker
- get weight euclidian, manhattan, etc
- write the A star function for baseline measurement
- write DQN algorithm
- write DDQN algorithm
- write DDPG etc
- pygame
- gym
The main game: morendod's repository