LEGO: Language Model Building Blocks

TL;DR

LEGO is a novel method for extracting and recombining small, task-specific language model blocks from large models, enhancing efficiency, privacy, and robustness in NLP applications.

Contribution

LEGO introduces a new approach combining LLM pruning, federated learning, and aggregation to create customizable, efficient, and privacy-preserving small language models.

Findings

LEGO enables efficient fine-tuning and inference with task-specific models.

LEGO maintains robustness and generalization despite data heterogeneity.

LEGO demonstrates versatility in model heterogeneity and privacy preservation.

Abstract

Large language models (LLMs) are essential in natural language processing (NLP) but are costly in data collection, pre-training, fine-tuning, and inference. Task-specific small language models (SLMs) offer a cheaper alternative but lack robustness and generalization. This paper proposes LEGO, a novel technique to extract SLMs from an LLM and recombine them. Using state-of-the-art LLM pruning strategies, we can create task- and user-specific SLM building blocks that are efficient for fine-tuning and inference while also preserving user data privacy. LEGO utilizes Federated Learning and a novel aggregation scheme for the LLM reconstruction, maintaining robustness without high costs and preserving user data privacy. We experimentally demonstrate the versatility of LEGO, showing its ability to enable model heterogeneity and mitigate the effects of data heterogeneity while maintaining LLM robustness.