yzk/veda-samhita-translation

Samhita Translation Dataset

Dataset Description

This dataset contains parallel sentence/verse-level pairs drawn from three Vedic Saṃhitā collections and translations:

• —R̥gveda Saṃhitā (rv)
• —Atharvaveda Saṃhitā (Śaunaka) (avs)
• —Taittirīya Saṃhitā (ts)

Each pair includes a Vedic Sanskrit source text and a reference translation. In the current release, the target language is English for avs and ts, and German for rv.

Dataset Structure

Data Instances

Each record contains the following fields:

• —text_id: unique identifier of the passage
• —source: source text in Vedic Sanskrit
• —target: reference translation
• —language: target language label

Licensing and Source Materials

This dataset is compiled from multiple external sources with different copyright and usage conditions.

• —rv ( R̥gveda): source text and translation were collected from VedaWeb.
• —avs (Atharvaveda Saṃhitā ): source text was collected from TITUS, and translations were collected from Wikisource.
• —ts (Taittirīya Saṃhitā ): source text was collected from TITUS, and translations were collected from the SanskritWeb-hosted PDF of Arthur Berriedale Keith's translation.

Because the dataset incorporates material from sources with different legal statuses and usage conditions, the dataset as a whole is not released under a single open license. This dataset is provided for non-commercial scholarly research only. Users must check and comply with the licensing and terms of use of the original source providers before redistributing, modifying, or using the data in downstream products.

Source Attribution

• —VedaWeb: https://vedaweb.uni-koeln.de/
• —TITUS: https://titus.uni-frankfurt.de/
• —Atharva-Veda Samhita on Wikisource: https://en.wikisource.org/wiki/Atharva-Veda_Samhita
• —SanskritWeb PDF of A. B. Keith's translation: http://sanskritweb.net/yajurveda/keith.pdf