Sparse Binary Compression: Towards Distributed Deep Learning with minimal Communication
May 22, 2018 Β· Declared Dead Β· π IEEE International Joint Conference on Neural Network
"No code URL or promise found in abstract"
Evidence collected by the PWNC Scanner
Authors
Felix Sattler, Simon Wiedemann, Klaus-Robert MΓΌller, Wojciech Samek
arXiv ID
1805.08768
Category
cs.LG: Machine Learning
Cross-listed
cs.AI,
cs.DC,
stat.ML
Citations
232
Venue
IEEE International Joint Conference on Neural Network
Last Checked
1 month ago
Abstract
Currently, progressively larger deep neural networks are trained on ever growing data corpora. As this trend is only going to increase in the future, distributed training schemes are becoming increasingly relevant. A major issue in distributed training is the limited communication bandwidth between contributing nodes or prohibitive communication cost in general. These challenges become even more pressing, as the number of computation nodes increases. To counteract this development we propose sparse binary compression (SBC), a compression framework that allows for a drastic reduction of communication cost for distributed training. SBC combines existing techniques of communication delay and gradient sparsification with a novel binarization method and optimal weight update encoding to push compression gains to new limits. By doing so, our method also allows us to smoothly trade-off gradient sparsity and temporal sparsity to adapt to the requirements of the learning task. Our experiments show, that SBC can reduce the upstream communication on a variety of convolutional and recurrent neural network architectures by more than four orders of magnitude without significantly harming the convergence speed in terms of forward-backward passes. For instance, we can train ResNet50 on ImageNet in the same number of iterations to the baseline accuracy, using $\times 3531$ less bits or train it to a $1\%$ lower accuracy using $\times 37208$ less bits. In the latter case, the total upstream communication required is cut from 125 terabytes to 3.35 gigabytes for every participating client.
Community Contributions
Found the code? Know the venue? Think something is wrong? Let us know!
π Similar Papers
In the same crypt β Machine Learning
R.I.P.
π»
Ghosted
R.I.P.
π»
Ghosted
XGBoost: A Scalable Tree Boosting System
R.I.P.
π»
Ghosted
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
R.I.P.
π»
Ghosted
Semi-Supervised Classification with Graph Convolutional Networks
R.I.P.
π»
Ghosted
Proximal Policy Optimization Algorithms
R.I.P.
π»
Ghosted
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Died the same way β π» Ghosted
R.I.P.
π»
Ghosted
Language Models are Few-Shot Learners
R.I.P.
π»
Ghosted
You Only Look Once: Unified, Real-Time Object Detection
R.I.P.
π»
Ghosted
A Unified Approach to Interpreting Model Predictions
R.I.P.
π»
Ghosted