MergeNet: Knowledge Migration across Heterogeneous Models, Tasks, and Modalities

TL;DR

MergeNet introduces a novel framework for transferring knowledge across diverse models, tasks, and modalities by learning to map parameter spaces, enabling effective knowledge transfer where traditional methods struggle.

Contribution

The paper proposes MergeNet, a method that bridges heterogeneous model parameter spaces for knowledge transfer, overcoming limitations of existing approaches that rely on shared structures.

Findings

Significant improvements in heterogeneous knowledge transfer tasks.

Effective transfer across different model architectures, tasks, and modalities.

Outperforms existing methods in challenging transfer scenarios.

Abstract

In this study, we focus on heterogeneous knowledge transfer across entirely different model architectures, tasks, and modalities. Existing knowledge transfer methods (e.g., backbone sharing, knowledge distillation) often hinge on shared elements within model structures or task-specific features/labels, limiting transfers to complex model types or tasks. To overcome these challenges, we present MergeNet, which learns to bridge the gap of parameter spaces of heterogeneous models, facilitating the direct interaction, extraction, and application of knowledge within these parameter spaces. The core mechanism of MergeNet lies in the parameter adapter, which operates by querying the source model's low-rank parameters and adeptly learning to identify and map parameters into the target model. MergeNet is learned alongside both models, allowing our framework to dynamically transfer and adapt knowledge relevant to the current stage, including the training trajectory knowledge of the source model. Extensive experiments on heterogeneous knowledge transfer demonstrate significant improvements in challenging settings, where representative approaches may falter or prove less applicable.