EBJR: Energy-Based Joint Reasoning for Adaptive Inference
This addresses the efficiency bottleneck for deploying AI models in real-world applications like cloud services, though it appears incremental as it builds on existing model combinations.
The paper tackles the problem of inefficient computational cost in state-of-the-art deep learning models by proposing an Energy-Based Joint Reasoning (EBJR) framework that adaptively distributes samples between large accurate models and small fast ones, achieving accuracy close to the deep model and latency close to the shallow one without requiring architecture changes or re-training.
State-of-the-art deep learning models have achieved significant performance levels on various benchmarks. However, the excellent performance comes at a cost of inefficient computational cost. Light-weight architectures, on the other hand, achieve moderate accuracies, but at a much more desirable latency. This paper presents a new method of jointly using the large accurate models together with the small fast ones. To this end, we propose an Energy-Based Joint Reasoning (EBJR) framework that adaptively distributes the samples between shallow and deep models to achieve an accuracy close to the deep model, but latency close to the shallow one. Our method is applicable to out-of-the-box pre-trained models as it does not require an architecture change nor re-training. Moreover, it is easy to use and deploy, especially for cloud services. Through a comprehensive set of experiments on different down-stream tasks, we show that our method outperforms strong state-of-the-art approaches with a considerable margin. In addition, we propose specialized EBJR, an extension of our method where we create a smaller specialized side model that performs the target task only partially, but yields an even higher accuracy and faster inference. We verify the strengths of our methods with both theoretical and experimental evaluations.