SHAKTI: A 2.5 Billion Parameter Small Language Model Optimized for Edge AI and Low-Resource Environments Paper • 2410.11331 • Published 17 days ago • 5