MLink: Linking Black-Box Models from Multiple Domains for Collaborative Inference

TL;DR

MLink introduces a novel approach to connect diverse black-box ML models through learned mappings, enabling cost-efficient multi-model inference with high accuracy in resource-constrained environments.

Contribution

The paper proposes a new model linking technique for heterogeneous black-box models and a scheduling algorithm, MLink, to improve inference efficiency under budget constraints.

Findings

MLink reduces inference computations by 66.7%

Achieves 94% inference accuracy under GPU memory constraints

Outperforms existing baselines in cost-efficiency and accuracy

Abstract

The cost efficiency of model inference is critical to real-world machine learning (ML) applications, especially for delay-sensitive tasks and resource-limited devices. A typical dilemma is: in order to provide complex intelligent services (e.g. smart city), we need inference results of multiple ML models, but the cost budget (e.g. GPU memory) is not enough to run all of them. In this work, we study underlying relationships among black-box ML models and propose a novel learning task: model linking, which aims to bridge the knowledge of different black-box models by learning mappings (dubbed model links) between their output spaces. We propose the design of model links which supports linking heterogeneous black-box ML models. Also, in order to address the distribution discrepancy challenge, we present adaptation and aggregation methods of model links. Based on our proposed model links, we developed a scheduling algorithm, named MLink. Through collaborative multi-model inference enabled by model links, MLink can improve the accuracy of obtained inference results under the cost budget. We evaluated MLink on a multi-modal dataset with seven different ML models and two real-world video analytics systems with six ML models and 3,264 hours of video. Experimental results show that our proposed model links can be effectively built among various black-box models. Under the budget of GPU memory, MLink can save 66.7% inference computations while preserving 94% inference accuracy, which outperforms multi-task learning, deep reinforcement learning-based scheduler and frame filtering baselines.