WitchesSocialStream/HowItsMade

---
task_categories:
- text-generation
- text-classification
language:
- en
pretty_name: How It's Made
size_categories:
- n<1K
configs:
- config_name: default
data_files:
- split: train
path: "HowItsMade.jsonl"
---

# Dataset Card for How It's Made

This tiny dataset contains parsed subtitles from the Canadian documentary series: "How It's Made".

## Dataset Details

### Dataset Description

- **Curated by:** KaraKaraWitch
- **Funded by [optional]:** N/A
- **Shared by [optional]:** N/A
- **Language(s) (NLP):** English
- **License:** Not Specified.

## Uses

This dataset is intended to be used in Large language models for grounding questions asking on "How X item is made?"

### Direct Use

N/A. Dataset released As-Is.

### Out-of-Scope Use

The auther thinks that this can be used to generate inaccurate descriptions.

## Dataset Structure

Refer to the image as shown


## Dataset Creation

### Curation Rationale

<!-- Motivation for the creation of this dataset. -->

[More Information Needed]

### Source Data

The true source is undisclosed. However, there are some more tech-savvy people who would know where this would come from. *wink.*

#### Data Collection and Processing

1. Obtain Episodes
2. Process the subtitles: Remove extra, split into sentences.

#### Who are the source data producers?

<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->

[More Information Needed]

## Bias, Risks, and Limitations

Weight and Temperatures uses US Imperial units instead of typical metric. Further processing is required for correct use.

### Recommendations

Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.

## Dataset Card Contact

Contact within the community tabs.

原始信息汇总

数据集概述：How Its Made

基本信息

• 任务类别：
  • 文本生成
  • 文本分类
• 语言： 英语
• 数据集大小： 小于1000条记录
• 配置：
  • 配置名称： default
  • 数据文件：
    • 分割： train
    • 路径： "HowItsMade.jsonl"

数据集描述

• 创建者： KaraKaraWitch
• 语言： 英语
• 许可证： 未指定

用途

• 主要用途： 用于大型语言模型中，回答关于“X物品是如何制造的？”的问题。
• 直接使用： 数据集按原样发布。
• 潜在风险： 可能被用于生成不准确的描述。

数据集结构

• 结构详情： 请参考提供的图片链接。

数据收集与处理

• 数据来源： 未公开，但技术娴熟的人可能知道其来源。
• 处理步骤：
  1. 获取剧集
  2. 处理字幕：去除多余部分，分割成句子。

偏差、风险和局限性

• 局限性： 使用美国英制单位而非公制单位，需要进一步处理以正确使用。

建议

• 用户注意事项： 应了解数据集的风险、偏差和局限性。