Accent Placement Models for Rigvedic Sanskrit Text

TL;DR

This work tackles automatic restoration of Rigvedic pitch accents by constructing a parallel accented/unaccented corpus and comparing three modeling approaches—full ByT5 fine-tuning, LoRA-based ByT5 tuning, and a BiLSTM-CRF baseline. It introduces robust evaluation using WER, CER, and a novel Diacritic Error Rate to isolate accent edits, showing that full ByT5 achieves the best overall fidelity while LoRA offers efficient performance and BiLSTM-CRF provides a transparent baseline. The results establish reproducible baselines for Rigvedic accent restoration and illuminate practical needs for Unicode-safe preprocessing, diacritic-aware tokenization, and application-ready metrics for OCR, ASR, and pedagogy. The study positions heritage-language NLP as a fruitful area bridging computational methods with philology and chanting pedagogy, with implications for future prosodic annotation and digitization efforts.

Abstract

The Rigveda, among the oldest Indian texts in Vedic Sanskrit, employs a distinctive pitch-accent system : udātta, anudātta, svarita whose marks encode melodic and interpretive cues but are often absent from modern e-texts. This work develops a parallel corpus of accented-unaccented ślokas and conducts a controlled comparison of three strategies for automatic accent placement in Rigvedic verse: (i) full fine-tuning of ByT5, a byte-level Transformer that operates directly on Unicode combining marks, (ii) a from-scratch BiLSTM-CRF sequence-labeling baseline, and (iii) LoRA-based parameter-efficient fine-tuning atop ByT5. Evaluation uses Word Error Rate (WER) and Character Error Rate (CER) for orthographic fidelity, plus a task-specific Diacritic Error Rate (DER) that isolates accent edits. Full ByT5 fine-tuning attains the lowest error across all metrics; LoRA offers strong efficiency-accuracy trade-offs, and BiLSTM-CRF serves as a transparent baseline. The study underscores practical requirements for accent restoration - Unicode-safe preprocessing, mark-aware tokenization, and evaluation that separates grapheme from accent errors - and positions heritage-language technology as an emerging NLP area connecting computational modeling with philological and pedagogical aims. Results establish reproducible baselines for Rigvedic accent restoration and provide guidance for downstream tasks such as accent-aware OCR, ASR/chant synthesis, and digital scholarship.