LorenaYannnnn/contrastive_instruction_tuning

---
license: apache-2.0
language:
- en
pretty_name: CoIN
size_categories:
- 10K<n<100K
---

# Dataset for Contrastive Instruction Tuning
This is the official dataset used in our ACL'24 Findings paper [Contrastive Instruction Tuning](https://arxiv.org/abs/2402.11138).

The original data source of our new dataset is the FLAN collection, specifically from [Muennighoff/flan](https://huggingface.co/datasets/Muennighoff/flan) on Huggingface. We follow the steps discussed in section 3.2 of the paper to curate dataset for CoIN.

- Each entry contains:
- The original instruction-input pair (```original_instruction```)
- The paraphrased instruction-input pair (```paraphrased_instruction```)
- Label (```targets```)
- Task name
- Keyword data (a dictionary that contains KV pairs that will be parsed into the instruction templates to get the full input).
- Every entry at the odd index is the hard negative for the entry above it.

# Github
The full implementation of our paper is available at [here](https://github.com/luka-group/CoIN?tab=readme-ov-file).

# Citation
```
@inproceedings{yan2024contrastive,
title={Contrastive Instruction Tuning},
author={Yan, Tianyi and Wang, Fei and Huang, James Y and Zhou, Wenxuan and Yin, Fan and Galstyan, Aram and Yin, Wenpeng and Chen, Muhao},
booktitle={ACL - Findings},
year={2024}
}
```

原始信息汇总

数据集概述

基本信息

• 许可证: Apache-2.0
• 语言: 英语 (en)
• 数据集大小: 10K<n<100K

数据集详情

• 名称: CoIN
• 用途: 用于ACL24 Findings论文《Contrastive Instruction Tuning》的官方数据集
• 原始数据来源: FLAN collection, 来自Huggingface的Muennighoff/flan

数据结构

• 每条记录包含:
  • 原始指令-输入对 (original_instruction)
  • 改写后的指令-输入对 (paraphrased_instruction)
  • 标签 (targets)
  • 任务名称
  • 关键词数据 (包含KV对的字典，用于解析指令模板以获取完整输入)
• 数据组织: 奇数索引的条目为其上方条目的困难负例

引用信息

@inproceedings{yan2024contrastive, title={Contrastive Instruction Tuning}, author={Yan, Tianyi and Wang, Fei and Huang, James Y and Zhou, Wenxuan and Yin, Fan and Galstyan, Aram and Yin, Wenpeng and Chen, Muhao}, booktitle={ACL - Findings}, year={2024} }