Gradient Accumulation
Gradient Accumulation is a computational technique in deep learning optimization whereby gradients are accumulated over multiple mini-batches before performing a single parameter update, enabling effective training with larger batch sizes under limited memory constraints.
23
Publications
3.6K
Citations
102
Authors
46
Institutions
Publications per year
2017–2026
16
Parents
23
Paulius Micikevicius, Sharan Narang, Jonah Alben et al. · arXiv (Cornell University) · 2017 · 881 citations · Full text
Paulius Micikevicius, Sharan Narang, Jonah Alben et al. · arXiv (Cornell University) · 2017 · 203 citations · Full text
1–5 of 23
102
2
h-index
First Affiliated Hospital of Xi'an Jiaotong University, China
2 publications · 50 citations
2
h-index
2
h-index
2
h-index
2
h-index
1–5 of 102
46

Beijing University of Posts and Telecommunications
Beijing, China
2 h-index
10 publications
705 citations

Carleton University
Ottawa, Canada
2 h-index
2 publications
141 citations
Tsinghua University
Beijing, China
2 h-index
9 publications
247 citations
Baidu (China)
Beijing, China
1 h-index
2 publications
406 citations

École Nationale Supérieure des Mines de Paris
Paris, France
1 h-index
2 publications
52 citations
Université de Montréal
Montreal, Canada
1 h-index
1 publications
1.8K citations

Brunel University of London
London, United Kingdom
1 h-index
1 publications
11 citations
Euphrosyne Polotskaya State University
Navapolatsk, Belarus
1 h-index
2 publications
22 citations
1–8 of 46
Repository · 5 publications · 1.2K citations
5
h-index
1 publications · 45 citations
1
h-index
1 publications · 30 citations
1
h-index
1 publications · 81 citations
1
h-index
1 publications · 91 citations
1
h-index
1–5