pytorch-hDQN

hierarchical-DQN in pytorch. [paper]

Results

Only Q-learning and h-DQN for StochasticMDPEnv are implemented.

Q-Learning for Experiment 1 (Discrete stochastic decision process)

h-DQN for Experiment 1 (Discrete stochastic decision process)

Reference

https://github.com/EthanMacdonald/h-DQN: Another implementation of hierarchical-DQN paper, which I copied the StochasticMDPEnv from.