ByteScale: Efficient Scaling of LLM Training with a 2048K Context Length on More Than 12,000 GPUs

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ge, Hao, Feng, Junda, Huang, Qi, Fu, Fangcheng, Nie, Xiaonan, Zuo, Lei, Lin, Haibin, Cui, Bin, Liu, Xin
Format:	Preprint
Published:	2025
Subjects:	Distributed, Parallel, and Cluster Computing Artificial Intelligence Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2502.21231