# CHARM✨ Benchmarking Chinese Commonsense Reasoning of LLMs: From Chinese-Specifics to Reasoning-Memorization Correlations [ACL2024]
[](https://arxiv.org/abs/2403.14112)
[](./LICENSE)
📃[Paper](https://arxiv.org/abs/2403.14112)
🏰[Project Page](https://opendatalab.github.io/CHARM/)
🏆[Leaderboard](https://opendatalab.github.io/CHARM/leaderboard.html)
✨[Findings](https://opendatalab.github.io/CHARM/findings.html)
## 数据集介绍
**CHARM** 是首个全面深入评估大型语言模型（LLMs）在中文常识推理能力的基准测试，它覆盖了国际普遍认知的常识以及独特的中国文化常识。此外，CHARM 还可以评估 LLMs 独立于记忆的推理能力，并分析其典型错误。
## 与其他常识推理评测基准的比较
## 🛠️ 如何使用
以下是快速下载 CHARM 并在 OpenCompass 上进行评估的步骤。
### 1. 下载 CHARM
```bash
git clone https://github.com/opendatalab/CHARM ${path_to_CHARM_repo}
```
### 2. 推理和评测
```bash
cd ${path_to_opencompass}
mkdir -p data
ln -snf ${path_to_CHARM_repo}/data/CHARM ./data/CHARM
# 在CHARM上对模型hf_llama3_8b_instruct做推理和评测
python run.py --models hf_llama3_8b_instruct --datasets charm_gen
```
## 🖊️ 引用
```bibtex
@misc{sun2024benchmarking,
title={Benchmarking Chinese Commonsense Reasoning of LLMs: From Chinese-Specifics to Reasoning-Memorization Correlations},
author={Jiaxing Sun and Weiquan Huang and Jiang Wu and Chenya Gu and Wei Li and Songyang Zhang and Hang Yan and Conghui He},
year={2024},
eprint={2403.14112},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```