Ask Your Humans: Using Human Instructions to Improve Generalization in Reinforcement Learning

This repository contains code that we used for data collection and for training the crafting agent. See subfolders for dataset collection code and training code.

Dataset

The full dataset can be downloaded here.

Please use this bibtex if you want to cite this work:

@inproceedings{chen2021ask,
title={Ask Your Humans: Using Human Instructions to Improve Generalization in Reinforcement Learning},
author={Valerie Chen and Abhinav Gupta and Kenneth Marino},
booktitle={International Conference on Learning Representations},
year={2021},
url={https://openreview.net/forum?id=Y87Ri-GNHYu}
}