silk-road/ChatHaruhi-English-62K-RolePlaying

---
license: cc-by-4.0
task_categories:
- text-generation
- text2text-generation
language:
- en
size_categories:
- 10K<n<100K
---

---
license: cc-by-4.0
task_categories:
- text-generation
- text2text-generation
language:
- zh
- en
size_categories:
- 100K<n<1M
---

# ChatHaruhi English_62K

20000 instance from original ChatHaruhi-54K
(translate original some chinese prompt into English)

42255 English Data from RoleLLM

token_len count via tokenizer from Phi-1.5

**github repo:**

https://github.com/LC1332/Chat-Haruhi-Suzumiya

Please star our github repo if you found the dataset is useful

## Regenerate Data

If you want to regenerate data with different context length, different embedding model or using your own chracter

now we refactored the final data generating pipeline

RoleLLM Data was generated by

https://github.com/LC1332/Haruhi-2-Dev/blob/main/notebook/%E4%BB%8EHF%E6%95%B0%E6%8D%AE%E6%A0%BC%E5%BC%8F%E5%BC%80%E5%A7%8B%E7%BB%84%E7%BB%87%E6%95%B0%E6%8D%AE(RoleLLM).ipynb

Chinese Data was generated by

https://github.com/LC1332/Haruhi-2-Dev/blob/main/notebook/%E9%A2%9D%E5%A4%96%E4%B8%AD%E6%96%87%E6%95%B0%E6%8D%AE.ipynb

63K Haruhi Data was generated by

https://github.com/LC1332/Haruhi-2-Dev/blob/main/notebook/%E4%BD%BF%E7%94%A8ChatHaruhi54K%E7%94%9F%E6%88%90%E6%95%B0%E6%8D%AE.ipynb

The data generated in speed around 8 instances per second under T4 GPU

The code was partly implemented by [豆角](https://github.com/goodnessSZW)

and finally verified by [LC1332](https://github.com/LC1332)

English Data was filtered by

https://github.com/LC1332/Haruhi-2-Dev/blob/main/notebook/Prepare_English_Training_set.ipynb

## Citation

Cite our paper if you use this dataset in your publication

```bibtex
@misc{li2023chatharuhi,
title={ChatHaruhi: Reviving Anime Character in Reality via Large Language Model},
author={Cheng Li and Ziang Leng and Chenxi Yan and Junyi Shen and Hao Wang and Weishi MI and Yaying Fei and Xiaoyang Feng and Song Yan and HaoSheng Wang and Linkang Zhan and Yaokai Jia and Pingyu Wu and Haozhen Sun},
year={2023},
eprint={2308.09597},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```

原始信息汇总

数据集概述

许可证信息

• 许可证类型: cc-by-4.0

任务分类

• 任务类别: 未提供具体分类信息