small language models

This article introduces LLaVA-Phi, a compact multi-modal assistant that utilizes the small language model, Phi-2, to facilitate intricate dialogues integrating both textual and visual elements. Despite having only 3 billion…