works on
distributed training 1fault-tolerant systems 1internet-scale compute 1large language models 1pipeline parallelism 1

From the 1 of 23 linked papers with an AI index.

activity
20242026
collaborators
Showing 2025 · cs.LGShow all

1 paper · 2 filters

cs.LG2025

Nesterov Method for Asynchronous Pipeline Parallel Optimization

Thalaiyasingam Ajanthan, Sameera Ramasinghe, Yan Zuo +2

Pipeline Parallelism (PP) enables large neural network training on small, interconnected devices by splitting the model into multiple stages. To maximize pipeline utilization, asyn…