1 paper
Shen Xu, Xiangwen Zhuge, Zhe Xu +3
LLMs often struggle with memory-constrained deployment on consumer-grade hardware due to their massive parameter sizes. While existing solutions such as model compression and offlo…