PARL-Sample/eight_puzzle at master · 11813006huanghao/PARL-Sample · GitHub

Name		Name	Last commit message	Last commit date
parent directory ..
imgs		imgs
learning_curve_log		learning_curve_log
naive_dqn		naive_dqn
rainbow		rainbow
DrawLearningCurve.py		DrawLearningCurve.py
EightPuzzleEnv.py		EightPuzzleEnv.py
Permutation.py		Permutation.py
README.md		README.md
ReversePairs.py		ReversePairs.py
StandardAStarSolve.py		StandardAStarSolve.py
TestCase.py		TestCase.py
__init__.py		__init__.py

README.md

整体架构:

对比Rainbow和普通DQN的平均奖励:

对比Rainbow和普通DQN的平均步数:

超参数影响:

对拍标准A*算法:

效果:

百度ai-studio博客地址: 当A*Star遇见PARL——使用Rainbow模型切ACM题