L-GreCo: Layerwise-Adaptive Gradient Compression for Efficient and Accurate Deep Learning

TL;DR

L-GreCo is a dynamic layerwise compression framework for distributed deep learning that improves training speed and compression efficiency without losing accuracy by adaptively tuning compression parameters per layer.

Contribution

It introduces a novel adaptive algorithm that optimally adjusts compression levels across layers during training, enhancing existing methods' effectiveness.

Findings

Achieves up to 2.5x training speedup.

Provides up to 5x compression improvement.

Maintains full model accuracy.

Abstract

Data-parallel distributed training of deep neural networks (DNN) has gained very widespread adoption, but can still experience communication bottlenecks. To address this issue, entire families of compression mechanisms have been developed, including quantization, sparsification, and low-rank approximation, some of which are seeing significant practical adoption. Despite this progress, almost all known compression schemes apply compression uniformly across DNN layers, although layers are heterogeneous in terms of parameter count and their impact on model accuracy. In this work, we provide a general framework for adapting the degree of compression across the model's layers dynamically during training, improving the overall compression, while leading to substantial speedups, without sacrificing accuracy. Our framework, called L-GreCo, is based on an adaptive algorithm, which automatically picks the optimal compression parameters for model layers guaranteeing the best compression ratio while satisfying an error constraint. Extensive experiments over image classification and language modeling tasks shows that L-GreCo is effective across all existing families of compression methods, and achieves up to 2.5$\times$ training speedup and up to 5$\times$ compression improvement over efficient implementations of existing approaches, while recovering full accuracy. Moreover, L-GreCo is complementary to existing adaptive algorithms, improving their compression ratio by 50% and practical throughput by 66%.