StableMoE: Stable Routing Strategy for Mixture of Experts

TL;DR

StableMoE introduces a two-stage training approach to stabilize routing in Mixture-of-Experts models, improving efficiency and performance in language tasks by reducing routing fluctuation issues.

Contribution

It proposes a novel two-stage training method that stabilizes expert routing in MoE models, enhancing convergence and task performance.

Findings

Outperforms existing MoE methods in convergence speed

Achieves better performance on language modeling tasks

Reduces routing fluctuation issues in MoE training

Abstract

The Mixture-of-Experts (MoE) technique can scale up the model size of Transformers with an affordable computational overhead. We point out that existing learning-to-route MoE methods suffer from the routing fluctuation issue, i.e., the target expert of the same input may change along with training, but only one expert will be activated for the input during inference. The routing fluctuation tends to harm sample efficiency because the same input updates different experts but only one is finally used. In this paper, we propose StableMoE with two training stages to address the routing fluctuation problem. In the first training stage, we learn a balanced and cohesive routing strategy and distill it into a lightweight router decoupled from the backbone model. In the second training stage, we utilize the distilled router to determine the token-to-expert assignment and freeze it for a stable routing strategy. We validate our method on language modeling and multilingual machine translation. The results show that StableMoE outperforms existing MoE methods in terms of both convergence speed and performance.