skip to main content


Title: Accordion: Adaptive Gradient Communication via Critical Learning Regime Identification
Distributed model training suffers from communication bottlenecks due to frequent model updates transmitted across compute nodes. To alleviate these bottlenecks, practitioners use gradient compression techniques like sparsification, quantization, or low-rank updates. The techniques usually require choosing a static compression ratio, often requiring users to balance the trade-off between model accuracy and per-iteration speedup. In this work, we show that such performance degradation due to choosing a high compression ratio is not fundamental. An adaptive compression strategy can reduce communication while maintaining final test accuracy. Inspired by recent findings on critical learning regimes, in which small gradient errors can have irrecoverable impact on model performance, we propose Accordion a simple yet effective adaptive compression algorithm. While Accordion maintains a high enough compression rate on average, it avoids over-compressing gradients whenever in critical learning regimes, detected by a simple gradient-norm based criterion. Our extensive experimental study over a number of machine learning tasks in distributed environments indicates that Accordion, maintains similar model accuracy to uncompressed training, yet achieves up to 5.5x better compression and up to 4.1x end-to-end speedup over static approaches. We show that Accordion also works for adjusting the batch size, another popular strategy for alleviating communication bottlenecks.  more » « less
Award ID(s):
1838733
NSF-PAR ID:
10299878
Author(s) / Creator(s):
; ; ; ;
Date Published:
Journal Name:
Fourth Conference on Machine Learning and Systems
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Smola, A. ; Dimakis, A. ; Stoica, I. (Ed.)
    Distributed model training suffers from communication bottlenecks due to frequent model updates transmitted across compute nodes. To alleviate these bottlenecks, practitioners use gradient compression techniques like sparsification, quantization, low rank updates etc. The techniques usually require choosing a static compression ratio, often requiring users to balance the trade-off between model accuracy and per-iteration speedup. In this work, we show that such performance degradation due to choosing a high compression ratio is not fundamental and that an adaptive compression strategy can reduce communication while maintaining final test accuracy.Inspired by recent findings on critical learning regimes, in which small gradient errors can have irrecoverable impact on model performance, we propose ACCORDION a simple yet effective adaptive compression algorithm. While ACCORDION maintains a high enough compression rate on average, it avoids detrimental impact by not compressing gradients too much whenever in critical learning regimes, detected by a simple gradient-norm based criterion. Our extensive experimental study over a number of machine learning tasks in distributed environments indicates that ACCORDION, maintains similar model accuracy to uncompressed training, yet achieves up to 5.5×better compression and up to 4.1×end-to-end speedup over static approaches. We show that ACCORDION also works for adjusting the batch size, another popular strategy for alleviating communication bottlenecks. Our code is available at https://github.com/uw-mad-dash/Accordion 
    more » « less
  2. null (Ed.)
    Deep learning (DL) is a popular technique for building models from large quantities of data such as pictures, videos, messages generated from edges devices at rapid pace all over the world. It is often infeasible to migrate large quantities of data from the edges to centralized data center(s) over WANs for training due to privacy, cost, and performance reasons. At the same time, training large DL models on edge devices is infeasible due to their limited resources. An attractive alternative for DL training distributed data is to use micro-clouds---small-scale clouds deployed near edge devices in multiple locations. However, micro-clouds present the challenges of both computation and network resource heterogeneity as well as dynamism. In this paper, we introduce DLion, a new and generic decentralized distributed DL system designed to address the key challenges in micro-cloud environments, in order to reduce overall training time and improve model accuracy. We present three key techniques in DLion: (1) Weighted dynamic batching to maximize data parallelism for dealing with heterogeneous and dynamic compute capacity, (2) Per-link prioritized gradient exchange to reduce communication overhead for model updates based on available network capacity, and (3) Direct knowledge transfer to improve model accuracy by merging the best performing model parameters. We build a prototype of DLion on top of TensorFlow and show that DLion achieves up to 4.2X speedup in an Amazon GPU cluster, and up to 2X speed up and 26% higher model accuracy in a CPU cluster over four state-of-the-art distributed DL systems. 
    more » « less
  3. Distributed machine learning is primarily motivated by the promise of increased computation power for accelerating training and mitigating privacy concerns. Unlike machine learning on a single device, distributed machine learning requires collaboration and communication among the devices. This creates several new challenges: (1) the heavy communication overhead can be a bottleneck that slows down the training, and (2) the unreliable communication and weaker control over the remote entities make the distributed system vulnerable to systematic failures and malicious attacks. This paper presents a variant of stochastic gradient descent (SGD) with improved communication efficiency and security in distributed environments. Our contributions include (1) a new technique called error reset to adapt both infrequent synchronization and message compression for communication reduction in both synchronous and asynchronous training, (2) new score-based approaches for validating the updates, and (3) integration with both error reset and score-based validation. The proposed system provides communication reduction, both synchronous and asynchronous training, Byzantine tolerance, and local privacy preservation. We evaluate our techniques both theoretically and empirically. 
    more » « less
  4. We explore a scheme that enables the training of a deep neural network in a Federated Learning configuration over an additive white Gaussian noise channel. The goal is to create a low complexity, linear compression strategy, called PolarAir, that reduces the size of the gradient at the user side to lower the number of channel uses needed to transmit it. The suggested approach belongs to the family of compressed sensing techniques, yet it constructs the sensing matrix and the recovery procedure using multiple access techniques. Simulations show that it can reduce the number of channel uses by ∼30% when compared to conveying the gradient without compression. The main advantage of the proposed scheme over other schemes in the literature is its low time complexity. We also investigate the behavior of gradient updates and the performance of PolarAir throughout the training process to obtain insight on how best to construct this compression scheme based on compressed sensing. 
    more » « less
  5. Federated Learning (FL) has attracted increasing attention in recent years. A leading training algorithm in FL is local SGD, which updates the model parameter on each worker and averages model parameters across different workers only once in a while. Although it has fewer communication rounds than the classical parallel SGD, local SGD still has large communication overhead in each communication round for large machine learning models, such as deep neural networks. To address this issue, we propose a new communicationefficient distributed SGD method, which can significantly reduce the communication cost by the error-compensated double compression mechanism. Under the non-convex setting, our theoretical results show that our approach has better communication complexity than existing methods and enjoys the same linear speedup regarding the number of workers as the full-precision local SGD. Moreover, we propose a communication-efficient distributed SGD with momentum, which also has better communication complexity than existing methods and enjoys a linear speedup with respect to the number of workers. At last, extensive experiments are conducted to verify the performance of our proposed methods. 
    more » « less