Proactive Gradient Conflict Mitigation in Multi-Task Learning: A Sparse Training Perspective

TL;DR

This paper introduces a sparse training approach to reduce gradient conflicts in multi-task learning, improving performance and compatibility with existing gradient manipulation methods.

Contribution

It proposes a novel sparse training strategy that mitigates gradient conflicts in multi-task learning, enhancing task performance and compatibility with other optimization techniques.

Findings

Sparse training reduces gradient conflicts effectively.

ST improves multi-task learning performance.

Compatible with existing gradient manipulation methods.

Abstract

Advancing towards generalist agents necessitates the concurrent processing of multiple tasks using a unified model, thereby underscoring the growing significance of simultaneous model training on multiple downstream tasks. A common issue in multi-task learning is the occurrence of gradient conflict, which leads to potential competition among different tasks during joint training. This competition often results in improvements in one task at the expense of deterioration in another. Although several optimization methods have been developed to address this issue by manipulating task gradients for better task balancing, they cannot decrease the incidence of gradient conflict. In this paper, we systematically investigate the occurrence of gradient conflict across different methods and propose a strategy to reduce such conflicts through sparse training (ST), wherein only a portion of the model's parameters are updated during training while keeping the rest unchanged. Our extensive experiments demonstrate that ST effectively mitigates conflicting gradients and leads to superior performance. Furthermore, ST can be easily integrated with gradient manipulation techniques, thus enhancing their effectiveness.