Comprehensive Open-Source Collection of Hindu Sacred Texts including Ramcharitmanas, Bhagavad Gita, Mahabharata, Valmiki Ramayana, Rigveda, Yajurveda, Atharvaveda for Educational and Research Purposes.

VedicGPT is an experimental transformer-based language model trained from scratch on a consolidated corpus of the Rigveda, Yajurveda, and Atharvaveda. The project explores how much structural and phonetic information can be learned from a relatively small Vedic Sanskrit dataset using modern NLP techniques.