Samas\=amayik: A Parallel Dataset for Hindi-Sanskrit Machine Translation

TL;DR

This paper introduces Samasamayik, a large-scale, diverse Hindi-Sanskrit parallel corpus, and demonstrates its effectiveness in improving machine translation performance for contemporary language data.

Contribution

The creation of a novel, large, and diverse Hindi-Sanskrit dataset covering modern sources, and benchmarking its utility with multiple translation models.

Findings

Models trained on Samasamayik outperform on in-domain data.

The dataset provides a strong baseline for Hindi-Sanskrit translation.

Minimal overlap with existing corpora confirms dataset's novelty.

Abstract

We release Samas\=amayik, a novel, meticulously curated, large-scale Hindi-Sanskrit corpus, comprising 92,196 parallel sentences. Unlike most data available in Sanskrit, which focuses on classical era text and poetry, this corpus aggregates data from diverse sources covering contemporary materials, including spoken tutorials, children's magazines, radio conversations, and instruction materials. We benchmark this new dataset by fine-tuning three complementary models - ByT5, NLLB and IndicTrans-v2, to demonstrate its utility. Our experiments demonstrate that models trained on the Samasamayik corpus achieve significant performance gains on in-domain test data, while achieving comparable performance on other widely used test sets, establishing a strong new performance baseline for contemporary Hindi-Sanskrit translation. Furthermore, a comparative analysis against existing corpora reveals minimal semantic and lexical overlap, confirming the novelty and non-redundancy of our dataset as a robust new resource for low-resource Indic language MT.