---
license: cc-by-nc-4.0
dataset_info:
  features:
  - name: text
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  splits:
  - name: train
    num_bytes: 3946452
    num_examples: 3067
  download_size: 1687475
  dataset_size: 3946452
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

# HerO at AVeriTeC
**[Paper](https://arxiv.org/abs/2410.12377) | [Code](https://github.com/ssu-humane/HerO)**

## Summaries
This is the modified instruction-tuning dataset derived from the [AVeriTeC dataset](https://huggingface.co/chenxwh/AVeriTeC), designed for fine-tuning large language models to predict verdicts and justifications in fact-checking tasks within AVeriTeC.

## License
This dataset is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License, following the original [AVeriTeC dataset](https://huggingface.co/chenxwh/AVeriTeC) license."

## Citation
```
@article{yoon2024hero,
  title={HerO at AVeriTeC: The Herd of Open Large Language Models for Verifying Real-World Claims},
  author={Yoon, Yejun and Jung, Jaeyoon and Yoon, Seunghyun and Park, Kunwoo},
  journal={arXiv preprint arXiv:2410.12377},
  year={2024}
}
```