Search papers, labs, and topics across Lattice.
The authors introduce SiPaKosa, a new corpus of 786K sentences (9.25M words) of Sinhala and Pali Buddhist texts, combining OCR'd historical manuscripts with web-scraped Tripitaka canonical texts. They perform rigorous quality control and metadata annotation, organizing the corpus into language-specific subcorpora. Benchmarking ten pretrained language models on SiPaKosa reveals that proprietary models achieve perplexity scores 3-6x lower than open-source alternatives.
Proprietary language models trounce open-source alternatives by 3-6x on a new, large-scale corpus of Sinhala and Pali Buddhist texts.
SiPaKosa is a comprehensive corpus of Sinhala and Pali doctrinal texts comprising approximately 786K sentences and 9.25M words, incorporating 16 copyright-cleared historical Buddhist documents alongside the complete web-scraped Tripitaka canonical texts. The corpus was created through high-quality OCR using Google Document AI on historical manuscripts, combined with systematic web scraping of canonical repositories, followed by rigorous quality control and metadata annotation. The corpus is organised into language-specific subcorpora: Sinhala and Mixed Sinhala-Pali. We evaluate the performance of language models using ten pretrained models, with perplexity scores ranging from 1.09 to 189.67 on our corpus. This analysis shows that proprietary models significantly outperform open-source alternatives by factors of three to six times. This corpus supports the pretraining of domain-adapted language models, facilitates historical language analysis, and aids in the development of information retrieval systems for Buddhist scholarship while preserving Sinhala cultural heritage.
