Lookahead: An Inference Acceleration Framework for Large Language Model with Lossless Generation Accuracy

Yao Zhao, Zhitian Xie, Chen Liang +2

As Large Language Models (LLMs) have made significant advancements across various tasks, such as question answering, translation, text summarization, and dialogue systems, the need…