When Precision Meets Position: BFloat16 Breaks Down RoPE in Long-Context Training

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Haonan, Liu, Qian, Du, Chao, Zhu, Tongyao, Du, Cunxiao, Kawaguchi, Kenji, Pang, Tianyu
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!