ResiHP: Taming LLM Training Failures with Dynamic Hybrid Parallelism

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ma, Tenghui, Guo, Jihu, Gao, Wei, Lu, Sitian, Ye, Zhisheng, Wang, Hanjing, Lin, Dahua
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!