Datasets

Published: 11 July 2023| Version 1 | DOI: 10.17632/f7b957zjyr.1
Contributor:
Lihua Zhang

Description

Expert demonstrations of Addressing Implicit Bias in Adversarial Imitation Learning, include: Hopper-v2, Walker2d-v2, HalfCheetah-v2 and Ant-v2

Files

Steps to reproduce

1. Configure the conda environment based on the requirement.txt or environment_variable.env file. 2. Unzip the expert demonstrations compressed file and the MI-GAIL compressed file. 3. Place the unzipped expert demonstrations folder into the unzipped MI-GAIL folder. 4. Run the main.py file. 5. The run_MI-GAIL.sh file is used for batch training.

Categories

Reinforcement Learning, Imitation Learning

Licence