Fetching the paper…

QuickLLaMA: Query-aware Inference Acceleration for Large Language Models · Around