DIORAMAS is a study in world-making, creating artificial, low-resolution worlds that mimic nature through pixel-based aesthetics and algorithmic growth patterns.
Introducing LLaVA-Phi: A Compact Vision-Language Assistant Powered By a Small Language Model | HackerNoon
LLaVA-Phi leverages the Phi-2 model to deliver effective multi-modal dialogues with only 2.7B parameters, demonstrating that smaller models can achieve high performance.