el2e10/aya-paraphrase-marathi

---
language:
- mr
license: cc
size_categories:
- n<1K
source_datasets:
- extended|ai4bharat/IndicXParaphrase
task_categories:
- text-generation
pretty_name: Aya Paraphrase Marathi
dataset_info:
features:
- name: inputs
dtype: string
- name: targets
dtype: string
- name: template_lang
dtype: string
- name: template_id
dtype: int64
splits:
- name: train
num_bytes: 683937
num_examples: 1001
download_size: 245473
dataset_size: 683937
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

### Description

This dataset is derived from the already existing dataset made by AI4Bharat. We have used the [IndicXParaphrase](https://huggingface.co/datasets/ai4bharat/IndicXParaphrase) dataset of AI4Bharat to create this instruction style dataset.
We have used the malayalam split of the above mentioned dataset to create this one. This was created as part of [Aya Open Science Initiative](https://sites.google.com/cohere.com/aya-en/home) from Cohere For AI.

IndicXParaphrase is multilingual, and n-way parallel dataset for paraphrase detection in 10 Indic languages. The original dataset(IndicXParaphrase) was made available under the cc-0 license.

### Template

The following templates(Marathi) where used for converting the original dataset:
```
#Template 1
prompt:
खालील वाक्य दुसरे-भिन्न शब्द वापरून लिहा: "{original_sentence}"

completion:
{paraphrased_sentence}
```
```
#Template 2
prompt:
खालील वाक्य वेगळ्या प्रकारे पुन्हा लिहा: "{original_sentence}"

completion:
{paraphrased_sentence}
```
```
#Template 3
prompt:
खालील वाक्य दुसरे शब्द वापरून रूपांतरित-अनुवादित करा: "{original_sentence}"

completion:
{paraphrased_sentence}
```

### Acknowledgement
Thank you, Yogesh Haribhau Kulkarni for helping with the preparation of this dataset by providing the Marathi translation of the above mentioned English prompts.

原始信息汇总

数据集概述

基本信息

• 语言: 马拉地语 (mr)
• 许可证: cc
• 数据集大小分类: n<1K
• 源数据集: 扩展自 ai4bharat/IndicXParaphrase
• 任务类别: 文本生成
• 数据集名称: Aya Paraphrase Marathi

数据集结构

• 特征:
  • inputs: 字符串类型
  • targets: 字符串类型
  • template_lang: 字符串类型
  • template_id: 整数类型 (int64)
• 分割:
  • train:
    • 字节数: 683937
    • 样本数: 1001
• 下载大小: 245473
• 数据集大小: 683937

配置

• 配置名称: default
  • 数据文件:
    • 分割: train
    • 路径: data/train-*

描述

该数据集源自AI4Bharat已有的IndicXParaphrase数据集，用于创建指令风格的马拉地语数据集。使用了IndicXParaphrase数据集的马拉雅拉姆语部分。

模板

以下模板用于转换原始数据集：

模板1

prompt: खालील वाक्य दुसरे-भिन्न शब्द वापरून लिहा: "{original_sentence}"

completion: {paraphrased_sentence}

模板2

prompt: खालील वाक्य वेगळ्या प्रकारे पुन्हा लिहा: "{original_sentence}"

completion: {paraphrased_sentence}

模板3

prompt: खालील वाक्य दुसरे शब्द वापरून रूपांतरित-अनुवादित करा: "{original_sentence}"

completion: {paraphrased_sentence}

致谢

感谢Yogesh Haribhau Kulkarni在准备此数据集时提供上述英文提示的马拉地语翻译。