Multi-task Hierarchical AIRL

Published:

GitHub Link

Codebase for my paper: Multi-task Hierarchical Adversarial Inverse Reinforcement Learning

Language: Python

The following parts are included:

  • Benchmarks built with Mujoco and D4RL, including Hopper, Walker, Ant, Kitchen, and Point maze.
  • An implementation of the multitask hierarchical imitation learning algorithm proposed in our paper and its ablations.
  • Implementations of the SOTA IL, HIL, and Multi-task (Meta) IL algorithms as baselines, including Option-GAIL, Directed-Info GAIL, MAML-BC, SMILE, PEMIRL.

To reproduce the baselines algorithms is quite challenging since I need to rewrite/reframe the codes for each of them to make them have the same coding style/structure as ours, in order to make the comparisons fair. At the same time, I need to stick to the pseudo-codes in their paper (my main reference), and also include the tricks and parameter setting used in their original implementations (if provided).