LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Models

Code for LLM-Planner.

Check project website for an overview and a demo.

News:

• Jun 24: Due to an object grounding error in the simulator we are using, we revert the code into high-level plan generation only before fix is implemented.

Quickstart

python hlp_planner.py

This commands uses the KNN dataset to generate a high-level plan for an example task. Check out the code for more details.

Hardware

Tested on:

• Mac M1
• Ubuntu 18.04

Citation Information

@InProceedings{song2023llmplanner,
  author    = {Song, Chan Hee and Wu, Jiaman and Washington, Clayton and Sadler, Brian M. and Chao, Wei-Lun and Su, Yu},
  title     = {LLM-Planner: Few-Shot Grounded Planning for Embodied Agents with Large Language Models},
  booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)},
  month     = {October},
  year      = {2023},
}

Acknowledgements

We thank the authors of ALFWORLD for releasing their code.

License

• LLM-Planner - MIT License
• ALFWorld - MIT License

Contact

Questions or issues? File an issue or contact Luke Song