CoChat

Sparse Communication for Distributed Gradient Descent

Open access
Aji, Alham Fikri, Heafield, Kenneth · 2017 · Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing

The DOI matches its Crossref record; 2 of the 2 references it cites were found in a registry or an index.

Abstract

We make distributed stochastic gradient descent faster by exchanging sparse updates instead of dense updates.Gradient updates are positively skewed as most updates are near zero, so we map the 99% smallest updates (by absolute value) to zero then exchange sparse matrices.This method can be combined with quantization to further improve the compression.We explore different configurations and apply them to neural machine translation and MNIST image classification tasks.Most configurations work on MNIST, whereas different configurations reduce convergence rate on the more complex translation task.Our experiments show that we can achieve up to 49% speed up on MNIST and 22% on NMT without damaging the final accuracy or BLEU.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.18653/v1/d17-1045

Cites

0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Distributed SGD architecture with parameter sharding.
Figure 1 Distributed SGD architecture with parameter sharding.CC-BY · From the PDF
Image 2
CC-BY · From the PDF
Figure 2 MNIST: Training loss and accuracy for different dropping ratios.
Figure 2 MNIST: Training loss and accuracy for different dropping ratios.CC-BY · From the PDF
Figure 4 MNIST: Comparison of local and global thresholds with and without layer normalization.
Figure 4 MNIST: Comparison of local and global thresholds with and without layer normalization.CC-BY · From the PDF
Figure 5 NMT: Comparison of local and global thresholds with and without layer normalization.
Figure 5 NMT: Comparison of local and global thresholds with and without layer normalization.CC-BY · From the PDF
Figure 6 MNIST classification accuracy over time.
Figure 6 MNIST classification accuracy over time.CC-BY · From the PDF
Figure 7 NMT validation BLEU and loss over time.
Figure 7 NMT validation BLEU and loss over time.CC-BY · From the PDF
Figure 8 Training loss for different quantization methods.
Figure 8 Training loss for different quantization methods.CC-BY · From the PDF

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

The publisher hasn’t deposited this paper’s reference list.

Research works printed but not linked: 16

  1. Tolu Alabi, Jeffrey D Blanchard, Bradley Gordon, and Russel Steinbach. 2012. Fast k-selection algorithms for graphics processing units. Journal of Experi- mental Algorithmics (JEA) 17:4-2.
  2. Dan Alistarh, Jerry Li, Ryota Tomioka, and Milan Vojnovic. 2016. QSGD: randomized quan- tization for communication-optimal stochas- tic gradient descent. CoRR abs/1610.02132. http://arxiv.org/abs/1610.02132. · arXiv
  3. Trishul M Chilimbi, Yutaka Suzue, Johnson Apacible, and Karthik Kalyanaraman. 2014. Project adam: Building an efficient and scalable deep learning training system. In OSDI. volume 14, pages 571- 582.
  4. Jeffrey Dean, Greg Corrado, Rajat Monga, Kai Chen, Matthieu Devin, Mark Mao, Andrew Senior, Paul Tucker, Ke Yang, Quoc V Le, et al. 2012. Large scale distributed deep networks. In Advances in neural information processing systems. pages 1223- 1231.
  5. Nikoli Dryden, Sam Ade Jacobs, Tim Moon, and Brian Van Essen. 2016. Communication quantization for data-parallel training of deep neural networks. In Proceedings of the Workshop on Machine Learn- ing in High Performance Computing Environments. IEEE Press, pages 1-8.
  6. Marcin Junczys-Dowmunt, Tomasz Dwojak, and Hieu Hoang. 2016. Is neural machine translation ready for deployment? A case study on 30 translation directions. In Program of the 13th International Workshop on Spoken Language Translation (IWSLT 2016).
  7. Yann LeCun, Léon Bottou, Yoshua Bengio, and Patrick Haffner. 1998. Gradient-based learning applied to document recognition. Proceedings of the IEEE 86(11):2278-2324.
  8. J. Lei Ba, J. R. Kiros, and G. E. Hinton. 2016. Layer Normalization. ArXiv e-prints .
  9. Ramón P Ñeco and Mikel L Forcada. 1996. Beyond mealy machines: Learning translators with recurrent neural networks. In Proceedings of the 1996 Inter- national Neural Network Society Annual Meeting. San Diego, California, USA.
  10. Kishore Papineni, Salim Roukos, Todd Ward, and Wei- Jing Zhu. 2002. BLEU: A method for automatic evalution of machine translation. In Proceedings 40th Annual Meeting of the Association for Com- putational Linguistics. Philadelphia, PA, pages 311- 318.
  11. Rajat Raina, Anand Madhavan, and Andrew Y Ng. 2009. Large-scale deep unsupervised learning using graphics processors. In Proceedings of the 26th an- nual international conference on machine learning. ACM, pages 873-880.
  12. Benjamin Recht, Christopher Re, Stephen Wright, and Feng Niu. 2011. Hogwild: A lock-free approach to parallelizing stochastic gradient de- scent. In J. Shawe-Taylor, R. S. Zemel, P. L. Bartlett, F. Pereira, and K. Q. Weinberger, editors, Advances in Neural Information Processing Sys- tems 24, Curran Associates, Inc., pages 693-701. http://papers.nips.cc/paper/4390-hogwild-a-lock- free-approach-to-parallelizing-stochastic-gradient- descent.pdf.
  13. Frank Seide, Hao Fu, Jasha Droppo, Gang Li, and Dong Yu. 2014. 1-bit stochastic gradi- ent descent and application to data-parallel distributed training of speech DNNs. In Interspeech. https://www.microsoft.com/en- us/research/publication/1-bit-stochastic-gradient- descent-and-application-to-data-parallel- distributed-training-of-speech-dnns/.
  14. Rico Sennrich, Barry Haddow, and Alexandra Birch. 2016. Edinburgh neural machine translation sys- tems for WMT 16. In Proceedings of the ACL 2016 First Conference on Machine Translation (WMT16).
  15. Martin Zinkevich, Markus Weimer, Lihong Li, and Alex J Smola. 2010. Parallelized stochastic gradient descent. In Advances in neural information process- ing systems. pages 2595-2603.
  16. Daniel Zwillinger and Stephen Kokoska. 1999. CRC standard probability and statistics tables and formu- lae. CRC Press.

Cited by

See all 783

Metadata

Authors
Aji, A. F., & Heafield, K.
Affiliations
University of Edinburgh (GB)
Venue
Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing
Published
2017
Pages
440-445
OpenAlex
W2606891064
Cited by
585
References
0
Citations by year
2026: 57 · 2025: 89 · 2024: 104 · 2023: 88 · 2022: 79 · 2021: 83 · 2020: 54 · 2019: 24 · 2018: 7
FWCI
10.60
Type
conference paper
Language
en
Open access
Licence
cc-by
Locations
  • Conference on Empirical Methods in Natural Language Processing (EMNLP) (publishedVersion, cc-by)
  • arXiv (Cornell University) (submittedVersion)
  • Edinburgh Research Explorer (University of Edinburgh) (submittedVersion, cc-by)
  • arXiv (Cornell University)
Funders
  • Nvidia
  • Alan Turing Institute
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Stochastic Gradient Optimization Techniques
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
Keywords
sparse communication, distributed gradient descent, stochastic gradient descent, gradient compression, quantization, neural machine translation, convergence rate, communication efficiency

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Sparse Communication for Distributed Gradient Descent. https://app.cochat.ai/papers/10.18653/v1/d17-1045
The paper
Aji, A. F., & Heafield, K. (2017). Sparse Communication for Distributed Gradient Descent. Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing. https://doi.org/10.18653/v1/d17-1045

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.