Sanskrit Grammar
MITRA Sanskrit Grammar
Advanced grammatical analysis for Sanskrit texts with Sandhi segmentation, lemmatization, and detailed morphological annotations available live at Dharmamitra.
We provide comprehensive grammatical analysis capabilities for Sanskrit powered by ByT5-Sanskrit model. This model represents the current state of the art for Sanskrit NLP, with an error rate roughly 50% lower than previous models and approaching the accuracy of a single human expert annotator.
Features
- Sandhi Segmentation: Automatic breaking down of compound words and Sandhi formations
- Lemmatization: Identification of base forms and dictionary entries for each word
- Grammatical Tags: Detailed morphological analysis including case, gender, number, tense, mood, and voice
- Lexical Candidates: Multiple possible meanings and interpretations for each word
- Interactive Interface: Click the 'grammar' button after entering Sanskrit text to access detailed annotations
How to Use
- Enter a Sanskrit sentence into the translation field at Dharmamitra
- Click the 'grammar' button that appears
- A side menu opens displaying comprehensive grammatical analysis including Sandhi segmentation, lemmatization, and grammatical tags
- Explore lexical candidates and morphological details for each word!
Technical Details
ByT5-Sanskrit is a grammatical annotation model trained on the Digital Corpus of Sanskrit by Oliver Hellwig