CAMEL: Communicative Agents for “Mind” Exploration of Large Scale Language Model Society

Description

• Github: https://github.com/lightaime/camel
• Website: https://www.camel-ai.org/
• Arxiv Paper: https://arxiv.org/abs/2303.17760

Dataset Summary

Physics dataset is composed of 20K problem-solution pairs obtained using gpt-4. The dataset problem-solutions pairs generating from 25 physics topics, 25 subtopics for each topic and 32 problems for each "topic,subtopic" pairs.

We provide the data in physics.zip.

Citation

@misc{li2023camel,
      title={CAMEL: Communicative Agents for "Mind" Exploration of Large Scale Language Model Society}, 
      author={Guohao Li and Hasan Abed Al Kader Hammoud and Hani Itani and Dmitrii Khizbullin and Bernard Ghanem},
      year={2023},
      eprint={2303.17760},
      archivePrefix={arXiv},
      primaryClass={cs.AI}
}

Prerequisite

None

License

cc-by-nc-4.0

Additional Information

Data Fields

The data fields for files in physics.zip are as follows:

• role_1: assistant role
• topic: physics topic
• sub_topic: physics subtopic belonging to topic
• message_1: refers to the problem the assistant is asked to solve.
• message_2: refers to the solution provided by the assistant.

Download in python

from huggingface_hub import hf_hub_download
hf_hub_download(repo_id="camel-ai/physics", repo_type="dataset", filename="physics.zip",
                local_dir="datasets/", local_dir_use_symlinks=False)