Chinese Evaluation Datasets
Collection
28 items • Updated • 5
Request access to ICE-PIXIU · NSP
This ICE-PIXIU task is released by The Fin AI for research. Access is granted automatically after you complete this short form.
By accessing this dataset you agree to the license of the original source (Public) and to cite the ICE-PIXIU paper and the original dataset in any resulting publication.
Log in or Sign Up to review the conditions and access this dataset content.
ICE-PIXIU · NSP
📄 Paper · 🌐 The Fin AI
Part of ICE-PIXIU — No Language is an Island: Unifying Chinese and English in Financial Large Language Models, Instruction Data, and Benchmarks (arXiv:2403.06249).
| Task | FinNJ |
| Original dataset | NSP (social texts) |
| Source license | Public |
| Language | zh |
Quick Start
from datasets import load_dataset
ds = load_dataset("TheFinAI/zh-nsp", split="test")
print(ds[0])
Example prompt
对于给出的金融新闻或社交媒体文本及其中包含的实体，你需要判断该文本中是否包含有针对所给实体的负面消息，只需回答"有"或者"无"。
Context: 【山东荣成处非办提示风险：恒昌汇财、纳觅财务等8家公司未提交备案材料】5月21日，有自媒体报道称，山东省威海市荣成市防范和处理非法集资工作领导小组办公室（下称荣成市处非办）于2018年5月18日发布风险提示，点名8家网贷机构未向荣成市金融监管（分享自@凤凰网财经）http://t.cn/R3kIaz5??
实体：恒昌汇财;纳觅财务
Answer:
Dataset Structure
| Split | Rows |
|---|---|
test |
500 |
| Field | Description |
|---|---|
choices |
Label space |
id |
Example id |
text |
Raw input text (without instruction) |
gold |
Index of the gold label in choices |
query |
Full instruction prompt given to the model |
answer |
Gold answer / label text |
License
The paper lists the original data as publicly available without a specific license (ICE-PIXIU paper, Table 1); refer to the original source for its terms.
Citation
Please cite ICE-PIXIU and the original dataset (NSP (social texts)):
@misc{hu2024languageislandunifyingchinese,
title={No Language is an Island: Unifying Chinese and English in Financial Large Language Models, Instruction Data, and Benchmarks},
author={Gang Hu and Ke Qin and Chenhan Yuan and Min Peng and Alejandro Lopez-Lira and Benyou Wang and Sophia Ananiadou and Jimin Huang and Qianqian Xie},
year={2024},
eprint={2403.06249},
archivePrefix={arXiv},
primaryClass={cs.CE},
url={https://arxiv.org/abs/2403.06249},
}
- Downloads last month
- 18