pharaouk/CoT-Collection

---
license: cc-by-4.0
task_categories:
- text-generation
- text-classification
language:
- en
size_categories:
- 1M<n<10M
---

# Dataset Card for Dataset Name

## Dataset Description

- **Homepage:https://github.com/kaistAI/CoT-Collection**
- **Repository:https://github.com/kaistAI/CoT-Collection**
- **Paper:https://arxiv.org/abs/2305.14045**
- **Point of Contact:[e-mail]**

### Dataset Summary


The CoT Collection is a dataset designed to induce Chain-of-Thought (CoT) capabilities into language models.
While proprietary LLMs excel at generating Chain-of-Thoughts based on prompting, smaller LMs do not have this capability. Thus, by fine-tuning to generate Chain-of-Thoughts, it could acquire such abilities.

The CoT Collection provides 1.84 million Chain-of-Thoughts augmented across 1060 tasks from the Flan Collection.\\
Experimental results show that fine-tuning on the CoT Collection results in (1) better zero-shot performance and (2) a better base model for few-shot learning.

We also provide a multilingual version of CoT Collection at this [link](https://huggingface.co/datasets/kaist-ai/Multilingual-CoT-Collection).

### Supported Tasks and Leaderboards

1060 tasks chosen from the Flan Collection.

The list of categories within the CoT Collection are:
* Natural Language Inference
* Extractive Question Answering
* Closed Book Question Answering
* Science
* Toxic Classification
* Arithmetic
* Program Execution
* Dialogue
* Ethics
* Commonsense Reasoning
* Multiple Choice Question Answering

### Languages

English

## Dataset Structure
* source: The input that is given to the language model (LM).
* target: The ground truth answer to the source.
* rationale: The Chain of Thought (CoT) that explains how the target could be derived from the source.
* task: A category that shows which dataset the source and target was extracted from.

In our paper, we trained the underlying language model to generate in the following format:
```
\{rationale\}
[RESULT]
\{target\}
```
Then during evaluation, we parsed the prediction after the phrase ```[RESULT]```.

### Data Splits

| name | train |
|-------------------|------:|
|CoT-Collection|1837928|

### Citation Information
If you find the following model helpful, please considering citing our paper!

```
@article{kim2023cot,
title={The CoT Collection: Improving Zero-shot and Few-shot Learning of Language Models via Chain-of-Thought Fine-Tuning},
author={Kim, Seungone and Joo, Se June and Kim, Doyoung and Jang, Joel and Ye, Seonghyeon and Shin, Jamin and Seo, Minjoon},
journal={arXiv preprint arXiv:2305.14045},
year={2023}
}
```

许可证：CC-BY-4.0
任务类别：
- 文本生成
- 文本分类
语言：
- 英语
规模区间：
- 100万<样本数<1000万
---

# 数据集卡片

## 数据集描述

- **项目主页：** https://github.com/kaistAI/CoT-Collection
- **代码仓库：** https://github.com/kaistAI/CoT-Collection
- **论文链接：** https://arxiv.org/abs/2305.14045
- **联络方式：** [e-mail]

### 数据集概述


CoT Collection是旨在为语言模型注入思维链（Chain-of-Thought，CoT）能力的数据集。专有大语言模型（Large Language Model，LLM）虽可凭借提示工程生成思维链，但小型LLM并不具备该能力。因此，通过微调使模型生成思维链，可让小型模型习得此类能力。

CoT Collection涵盖源自Flan Collection的1060项任务，共计184万条思维链数据。实验结果表明，基于CoT Collection进行微调可实现两项优势：(1) 更优异的零样本（Zero-shot）性能；(2) 适用于少样本（Few-shot）学习的更优质基础模型。

此外，我们在该[链接](https://huggingface.co/datasets/kaist-ai/Multilingual-CoT-Collection)中提供了CoT Collection的多语言版本。

### 支持任务与评测基准

从Flan Collection中遴选的1060项任务。

CoT Collection涵盖的任务类别如下：
* 自然语言推理
* 抽取式问答
* 闭卷问答
* 科学任务
* 毒性分类
* 算术任务
* 程序执行
* 对话
* 伦理判断
* 常识推理
* 多项选择题问答

### 语言

英语

## 数据集结构

* **source**：输入至语言模型（Language Model，LM）的待处理文本。
* **target**：对应source的标准答案。
* **rationale**：用于阐释如何从source推导得到target的思维链（Chain-of-Thought，CoT）。
* **task**：用于标识source与target所属数据集类别的标签。

在我们的论文中，我们将基础语言模型的训练格式设定为：

{rationale}
[RESULT]
{target}

在评估阶段，我们会提取`[RESULT]`之后的内容作为模型的预测输出。

### 数据划分

| 数据集名称 | 训练集样本数 |
|-------------------|------:|
|CoT-Collection|1837928|

### 引用信息

如果您认为本数据集对您的研究有所帮助，请引用我们的论文！

@article{kim2023cot,
title={The CoT Collection: Improving Zero-shot and Few-shot Learning of Language Models via Chain-of-Thought Fine-Tuning},
author={Kim, Seungone and Joo, Se June and Kim, Doyoung and Jang, Joel and Ye, Seonghyeon and Shin, Jamin and Seo, Minjoon},
journal={arXiv preprint arXiv:2305.14045},
year={2023}
}

原始信息汇总

数据集概述

数据集名称

• 名称: CoT Collection

数据集描述

• 目的: 旨在通过微调使语言模型具备Chain-of-Thought (CoT)能力。
• 规模: 包含1.84百万个CoT，覆盖1060个任务。
• 效果: 微调后，模型在零样本学习和少样本学习中表现更佳。

支持的任务

• 任务数量: 1060个任务
• 任务类型:
  • 自然语言推理
  • 提取式问答
  • 闭卷问答
  • 科学
  • 毒性分类
  • 算术
  • 程序执行
  • 对话
  • 伦理
  • 常识推理
  • 多选题问答

语言

• 主要语言: 英语

数据集结构

• 数据字段:
  • source: 输入给语言模型的数据
  • target: 源数据的真实答案
  • rationale: 解释如何从源数据推导出目标答案的CoT
  • task: 显示源数据和目标数据来自哪个数据集的类别

数据分割

• 分割:
  • 训练集: 1837928条数据

引用信息

• 引用文献:

  @article{kim2023cot, title={The CoT Collection: Improving Zero-shot and Few-shot Learning of Language Models via Chain-of-Thought Fine-Tuning}, author={Kim, Seungone and Joo, Se June and Kim, Doyoung and Jang, Joel and Ye, Seonghyeon and Shin, Jamin and Seo, Minjoon}, journal={arXiv preprint arXiv:2305.14045}, year={2023} }