FedMKT: Federated Mutual Knowledge Transfer for Large and Small Language Models

TL;DR

FedMKT introduces a federated framework that simultaneously enhances large language models and small language models through mutual knowledge transfer, improving NLP performance in a privacy-preserving, efficient manner.

Contribution

This paper presents FedMKT, a novel federated mutual knowledge transfer method that jointly improves large and small language models, addressing a key gap in existing federated learning approaches.

Findings

FedMKT significantly improves both LLM and SLM performance.

The framework effectively transfers knowledge using token alignment with MinED.

Experimental results across multiple scenarios validate the approach's effectiveness.

Abstract

Recent research in federated large language models (LLMs) has primarily focused on enabling clients to fine-tune their locally deployed homogeneous LLMs collaboratively or on transferring knowledge from server-based LLMs to small language models (SLMs) at downstream clients. However, a significant gap remains in the simultaneous mutual enhancement of both the server's LLM and clients' SLMs. To bridge this gap, we propose FedMKT, a parameter-efficient federated mutual knowledge transfer framework for large and small language models. This framework is designed to adaptively transfer knowledge from the server's LLM to clients' SLMs while concurrently enriching the LLM with clients' unique domain insights. We facilitate token alignment using minimum edit distance (MinED) and then selective mutual knowledge transfer between client-side SLMs and a server-side LLM, aiming to collectively enhance their performance. Through extensive experiments across three distinct scenarios, we evaluate the effectiveness of FedMKT using various public LLMs and SLMs on a range of NLP text generation tasks. Empirical results demonstrate that FedMKT simultaneously boosts the performance of both LLMs and SLMs.