chituHigh-performance inference framework for large language models, focusing on efficiency, flexibility, and availability.ATK2926DEF2056