the-ak-2000/vedic-texts

Vedic Texts Dataset

Dataset Description

• —Repository: the-ak-2000/vedic-texts
• —Paper: [If applicable]
• —Point of Contact: [Your contact information]

Dataset Summary

This dataset is a comprehensive collection of Vedic texts from various sources, including:

• —Rigveda (Samhita)
• —Sankhayana Brahmana
• —Satapatha Brahmana
• —Upanishads
• —Vedangas

The dataset contains 65,140 verses in total, with 58,626 verses in the training set and 6,514 verses in the test set. Each verse is annotated with metadata including its source, genre, and text-specific structural information.

Supported Tasks and Leaderboards

This dataset can be used for:

• —Sanskrit text analysis
• —Vedic literature research
• —Historical linguistics
• —Religious studies
• —Natural Language Processing tasks in Sanskrit

Dataset Structure

Data Fields

• —text (string): The actual verse text in Sanskrit
• —source (string): Path to the source file
• —metadata (dictionary): A structured dictionary containing text-specific metadata:
• —text_type (string): Type of text (samhita, brahmana, upanishad, vedanga)
• —genre (string): Genre of the text
• —book (string): Book name/number

For Samhita texts (e.g., Rigveda):

• —mandala (integer): Book number
• —sukta (integer): Hymn number
• —verse (integer): Verse number
• —line (integer): Line number
• —varga (integer): Section number

For Brahmana texts (e.g., Satapatha):

• —kanda (integer): Book number
• —adhyaya (integer): Chapter number
• —brahmana (integer): Brahmana number
• —kandika (integer): Section number

For Upanishads:

• —upanishad (string): Name of the Upanishad
• —chapter (integer): Chapter number
• —section (integer): Section number
• —verse (integer): Verse number

For Vedangas:

• —commentary_type (string): Type of commentary
• —file (string): Original file name
• —sutra_number (string): Sutra number

Data Splits

• —Training set: 58,626 verses
• —Test set: 6,514 verses

Source Data

Initial Data Collection and Normalization

The texts were collected from various sources and normalized to ensure consistent formatting and encoding. The dataset includes:

• —Rigveda: The oldest layer of Vedic texts
• —Sankhayana Brahmana: A commentary on the Rigveda
• —Satapatha Brahmana: A detailed commentary on the Yajurveda
• —Upanishads: Philosophical texts that form the basis of Vedanta
• —Vedangas: Auxiliary disciplines of Vedic study

Who are the source language producers?

The texts were originally composed by ancient Vedic seers (rishis) and preserved through oral tradition before being written down. The dataset includes texts from various periods of Vedic literature, spanning from approximately 1500 BCE to 500 BCE.

Personal and Sensitive Information

This dataset contains religious and philosophical texts that are part of the public domain. No personal or sensitive information is included.

Dataset Creation

Curation Rationale

The dataset was created to provide a comprehensive collection of Vedic texts for research and analysis purposes. The texts were selected to represent different genres and periods of Vedic literature.

Source Language

The texts are in Sanskrit, the classical language of ancient India.

Annotations

The dataset includes structural annotations specific to each text type, preserved in the metadata field. This allows for flexible querying and analysis while maintaining the original structure of each text.

Personal and Sensitive Information

No personal or sensitive information is included in the dataset.

Additional Information

Dataset Curators

[Your name/organization]

Licensing Information

The texts in this dataset are in the public domain. The dataset itself is released under the Creative Commons Attribution 4.0 International License.

Citation Information

If you use this dataset, please cite it as:

@misc{vedic_texts_dataset,
  author = {[Your name]},
  title = {Vedic Texts Dataset},
  year = {2024},
  publisher = {Hugging Face},
  journal = {Hugging Face Hub},
  howpublished = {\url{https://huggingface.co/datasets/the-ak-2000/vedic-texts}}
}

Contributions

If you'd like to contribute to this dataset, please contact the maintainers.

Dataset Card Authors

[Your name/organization]