Asking for Knowledge : Training RL Agents to Query External Knowledge Using Language
[Project Website]
The repository contains Pytorch implementation of the Q-BabyAI, Q-TextWorld environments, and Asking for knowledge (AFK) agent.
If you used this code for your experiments or found it helpful, please consider citing the following paper:
@inproceedings{AFK2022, author = {Iou-Jen Liu$^\ast$ and Xingdi Yuan$^\ast$ and Marc-Alexandre C\^{o}t\'{e}$^\ast$ and Pierre-Yves Oudeyer and Alexander G. Schwing}, title = {Asking for Knowledge: Training RL Agents to Query External Knowledge Using Language}, booktitle = {arXiv.}, year = {2022}, note = {$^\ast$ equal contribution} }
Please see README.md in afk-q-babyai
and afk-q-textworld/text_agent
folder for detailed instruction for running the code.
Acknowledgement
The Q-babyAI environments and the RL agent are based on the BabyAI of https://github.com/mila-iqia/babyai
License
AFK is licensed under the MIT License