SC21 AD/AE for "Why Globally Re-shuffle? An I/O Perspective on Data Shuffling in Large Scale Deep Learning"

Mohamed WahibFigshare 2021 · DOI · OpenAlex

Abstract

We use ABCI (https://abci.ai/) for testing the effect of shuffling techniques on convergence of distributed SGD training<br> This repository contains the artifact, e.g. job scripts, models, datasets.
BibTeX
@inproceedings{wahib2021sc,
  author = {Mohamed Wahib},
  title = {{SC21 AD/AE for "Why Globally Re-shuffle? An I/O Perspective on Data Shuffling in Large Scale Deep Learning"}},
  booktitle = {Figshare},
  year = {2021},
  doi = {10.5281/zenodo.4678274},
}

Loading related content...

Figure