Quandong Wang

SUBLLM: A Novel Efficient Architecture with Token Sequence Subsampling for LLM

2024 · arXiv (Cornell University)

While Large Language Models (LLMs) have achieved remarkable success in various fields, the efficiency of training and inference remains a major challenge. To address this issue, we propose SUBLLM, short for Subsampling-Upsampling-Bypass Large Language Model, …