Code associated with our work: "Sample-Efficient Imitation Learning via Generative Adversarial Nets"
Published in AISTATS 2019 | arXiv link | Video demos | Expert demonstrations
Launching scripts are available in /launchers.
- Interaction simulated with the openai/gym API.
- Some utilities were inspired from openai/baselines.