Compressed Communication for Distributed Training: Adaptive Methods and System

by   Yuchen Zhong, et al.

Communication overhead severely hinders the scalability of distributed machine learning systems. Recently, there has been a growing interest in using gradient compression to reduce the communication overhead of the distributed training. However, there is little understanding of applying gradient compression to adaptive gradient methods. Moreover, its performance benefits are often limited by the non-negligible compression overhead. In this paper, we first introduce a novel adaptive gradient method with gradient compression. We show that the proposed method has a convergence rate of 𝒪(1/√(T)) for non-convex problems. In addition, we develop a scalable system called BytePS-Compress for two-way compression, where the gradients are compressed in both directions between workers and parameter servers. BytePS-Compress pipelines the compression and decompression on CPUs and achieves a high degree of parallelism. Empirical evaluations show that we improve the training time of ResNet50, VGG16, and BERT-base by 5.0 23.3 Furthermore, for training the BERT models, we achieve a compression rate of 333x compared to the mixed-precision training.


page 1

page 2

page 3

page 4


MergeComp: A Compression Scheduler for Scalable Communication-Efficient Distributed Training

Large-scale distributed training is increasingly becoming communication ...

Fine-tuning Language Models over Slow Networks using Activation Compression with Guarantees

Communication compression is a crucial technique for modern distributed ...

ByteComp: Revisiting Gradient Compression in Distributed Training

Gradient compression (GC) is a promising approach to addressing the comm...

Training Faster with Compressed Gradient

Although the distributed machine learning methods show the potential for...

1-bit LAMB: Communication Efficient Large-Scale Large-Batch Training with LAMB's Convergence Speed

To train large models (like BERT and GPT-3) with hundreds or even thousa...

Is Network the Bottleneck of Distributed Training?

Recently there has been a surge of research on improving the communicati...

Trajectory Normalized Gradients for Distributed Optimization

Recently, researchers proposed various low-precision gradient compressio...

Please sign up or login with your details

Forgot password? Click here to reset