notesum.ai

Published at May 9

Towards Next-Level Post-Training Quantization of Hyper-Scale Transformers

NeurIPS

Released Date: May 9, 2024

Authors: Junhan Kim1, Chungman Lee1, Eulrang Cho1, Kyungphil Park1, Ho-young Kim1, Joonyoung Kim1, Yongkweon Jeon1

Aff.: 1Samsung Research

Arxiv: https://openreview.net/pdf/92b693f63079c7ab2c8fcfce715dcc49a20e6fd4.pdf