Textcraft Environment

This is the Pypi implementation and packaging of the environment Textcraft introduced in the paper ADaPT.

Getting started:

The package comes with all the data already, so usage is super easy.

1. Installation:

pip3 install textcraft

2. Usage:

from textcraft import TextCraft


env = TextCraft()
obs, info = env.reset(seed=42)
print(obs)
action = input("> ")
(observation, reward, terminated, truncated, info) = env.step(action)
print(observation, reward, sep="\n")

Acknowledgements

We thank the authors and contributors of ADaPT for their public code release.

Reference

Please cite StateAct:

@article{rozanov2024stateactstatetrackingreasoning,
      title={StateAct: State Tracking and Reasoning for Acting and Planning with Large Language Models}, 
      author={Nikolai Rozanov and Marek Rei},
      year={2024},
      eprint={2410.02810},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2410.02810}, 
}

ADaPT

@article{prasad2023adapt,
      author    = "Prasad, Archiki and Koller, Alexander and Hartmann, Mareike and Clark, Peter and Sabharwal, Ashish and Bansal, Mohit and Khot, Tushar",
      title     = "ADaPT: As-Needed Decomposition and Planning with Language Models",
      journal   = "arXiv",
      year      = "2023",}
}