Emergent Mind

Compressing Large Language Models by Streamlining the Unimportant Layer

(2403.19135)
Published Mar 28, 2024 in cs.CL and cs.AI

Abstract

Large language models (LLM) have been extensively applied in various natural language tasks and domains, but their applicability is constrained by the large number of parameters of the models. Consequently, there is an increasing emphasis on compact models that exhibit high performance. In this study, we observe that different layers in LLM have varying degrees of perturbation on the hidden states, which allows us to identify less important layers. Based on this phenomenon, we propose LLM-Streamline, which consists of two parts: layer pruning, where we remove a set of consecutive layers with the lowest importance in the model according to the target sparsity; and layer replacement, where we train a lightweight model to substitute the pruned layers, thereby mitigating the performance degradation caused by pruning. In our experiments, we utilize structures such as a multi-layer perceptron (MLP) and a transformer layer as lightweight models and ultimately demonstrate that a single MLP can effectively fit the pruned layers. Comprehensive experiments show that our proposed method, LLM-Streamline, outperforms previous state-of-the-art (SOTA) model pruning methods.

We're not able to analyze this paper right now due to high demand.

Please check back later (sorry!).

Generate a summary of this paper on our Pro plan:

We ran into a problem analyzing this paper.

Newsletter

Get summaries of trending comp sci papers delivered straight to your inbox:

Unsubscribe anytime.