nnScaler-M: Constraint-Guided and Placement-Aware Parallelization Plan Generation for Deep Learning Training
- 1 July 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Parallel and Distributed Systems
- Vol. 37 (7) , 1691-1708
Abstract
No abstract availableKeywords
Funding Information
- National Key R&D Program of China (2024YFB4505701)
- National Natural Science Foundation of China (62572454)
This publication has 19 references indexed in Scilit:
- Tessel: Boosting Distributed Execution of Large DNN Models via Flexible Schedule Search2024
- Swin Transformer V2: Scaling Up Capacity and Resolution2022
- Highly accurate protein structure prediction with AlphaFoldNature, 2021
- ZeRO: Memory optimizations Toward Training Trillion Parameter Models2020
- DeepSpeed2020
- SwapAdvisor2020
- PipeDream2019
- A generic communication scheduler for distributed DNN training acceleration2019
- Supporting Very Large Models using Automatic Dataflow Graph Partitioning2019
- Ako2016