Degree-Quant: Quantization-Aware Training for Graph Neural Networks
This work addresses the efficiency problem for deploying GNNs in resource-constrained environments, representing an incremental advancement in quantization methods for a specific domain.
The paper tackles the problem of making graph neural networks (GNNs) more efficient at inference time by developing a quantization-aware training method called Degree-Quant, which identifies unique quantization errors in GNNs and improves performance over baselines, achieving up to 26% gains for INT4 models and enabling up to 4.7x speedups on CPU with INT8 arithmetic.
Graph neural networks (GNNs) have demonstrated strong performance on a wide variety of tasks due to their ability to model non-uniform structured data. Despite their promise, there exists little research exploring methods to make them more efficient at inference time. In this work, we explore the viability of training quantized GNNs, enabling the usage of low precision integer arithmetic during inference. We identify the sources of error that uniquely arise when attempting to quantize GNNs, and propose an architecturally-agnostic method, Degree-Quant, to improve performance over existing quantization-aware training baselines commonly used on other architectures, such as CNNs. We validate our method on six datasets and show, unlike previous attempts, that models generalize to unseen graphs. Models trained with Degree-Quant for INT8 quantization perform as well as FP32 models in most cases; for INT4 models, we obtain up to 26% gains over the baselines. Our work enables up to 4.7x speedups on CPU when using INT8 arithmetic.