1 paper
Li Zhang, Youhe Jiang, Guoliang He +6
Mixed-precision inference techniques reduce the memory and computational demands of Large Language Models (LLMs) by applying hybrid precision formats to model weights, activations,…