An open-source bilingual Kannada-English Large Language Model designed for language adaptability and trained on a compact dataset.

A training phase that enhances language understanding by alternating sentences between Kannada and English for next-token prediction.