Datasets:

keplersystems
/

PAARI-Marathi-TTS

language:
  - mr
tags:
  - journalism
  - rural-india
  - paari
  - mr
  - devanagari
  - tts
  - speech-synthesis
size_categories:
  - n<1K
task_categories:
  - text-to-speech
  - text-classification
multilinguality:
  - monolingual
pretty_name: PAARI Marathi TTS Dataset
dataset_info:
  features:
    - name: article_id
      dtype: int64
    - name: chunk_id
      dtype: int64
    - name: title
      dtype: string
    - name: text
      dtype: string
    - name: text_length
      dtype: int64
    - name: total_chunks
      dtype: int64
    - name: language
      dtype: string
    - name: language_name
      dtype: string
    - name: audio
      dtype: audio
  splits:
    - name: train
      num_bytes: 22238984435.72
      num_examples: 9956
  download_size: 21521940916
  dataset_size: 22238984435.72
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

PAARI Marathi TTS Dataset

Dataset Description

This dataset contains TTS-optimized chunks of journalism articles in Marathi (मराठी) from the People's Archive of Rural India (PAARI). The articles focus on rural life, agriculture, social issues, and cultural stories from rural India.

Dataset Details

Language: Marathi (मराठी)
Script: Devanagari
Language Code: mr
Dataset Type: TTS-optimized
Source: Rural India Online
License: Please refer to PAARI's terms of use

Dataset Structure

TTS Dataset Fields

article_id: Unique identifier for the source article
chunk_id: Sequential chunk number within the article
title: Article title in Marathi
text: Chunked article content optimized for TTS (≤800 characters) in Marathi
text_length: Length of the text chunk in characters
total_chunks: Total number of chunks for this article
language: Language code
language_name: Full language name

Usage

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("keplersystems/PAARI-Marathi-TTS")

# Access examples
for chunk in dataset["train"]:
    print(f"Article {chunk["article_id"]}, Chunk {chunk["chunk_id"]}: {chunk["text"][:100]}...")

Text Processing

The text has been cleaned and processed with:

HTML entity decoding
HTML tag removal
Devanagari-specific text normalization
Whitespace normalization
Sentence-boundary aware chunking for TTS optimization

Data Quality

Articles filtered for completeness (both title and text required)
Language-specific text cleaning applied
Optimal chunk sizes for speech synthesis

Citation

If you use this dataset, please cite:

@misc{paari-mr-tts-dataset,
    title={PAARI Marathi TTS Dataset},
    author={People's Archive of Rural India},
    howpublished={\url{https://huggingface.co/datasets/keplersystems/PAARI-Marathi-TTS}},
    year={2025}
}

Acknowledgments

This dataset is derived from the excellent journalism work of the People's Archive of Rural India (PAARI) and Rural India Online. We thank all the journalists, photographers, and contributors who created this valuable content documenting rural Indian life and culture.

Contact

For questions about this dataset, please open an issue on the dataset repository.