3 papers
cs.DC2025
Redox: Improving I/O Efficiency of Model Training Through File Redirection
Yuhao Li, Xuanhua Shi, Yunfei Zhao +3
This paper proposes Redox, a training data management system designed to achieve high I/O efficiency. The key insight is a new observation of file redirection: for model training,…
cs.DC2025
CFP: Efficient Optimization of Intra-Operator Parallelism Plans for Large Model Training
Weifang Hu, Xuanhua Shi, Yunkai Zhang +7
Optimizing the parallel training of large models requires exploring intra-operator parallelism plans for a computation graph that typically contains tens of thousands of primitive…
cs.DC2025
DawnPiper: A Memory-scablable Pipeline Parallel Training Framework
Xuan Peng, Xuanhua Shi, Haolin Zhang +2
Pipeline parallelism is a crucial paradigm for large-scale model training. However, imbalances in memory footprint across stages can lead to significant GPU memory wastage, limitin…