Effective Distillation of Table-based Reasoning Ability from LLMs

TL;DR

This paper introduces a novel distillation method to transfer table-based reasoning skills from large language models to smaller models, significantly improving their performance on scientific table-to-text tasks.

Contribution

The paper presents a new distillation approach specifically for table reasoning, enabling smaller models to outperform larger LLMs on scientific table-to-text generation.

Findings

Smaller models fine-tuned with distilled data outperform traditional fine-tuning.

Distilled models surpass certain LLMs on scientific table-to-text datasets.

The approach enhances table reasoning capabilities in resource-efficient models.

Abstract

Large Language Models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing tasks. However, their enormous parameter size and extremely high requirements for compute power pose challenges for their practical deployment. Recent research has revealed that specific capabilities of LLMs, such as numerical reasoning, can be transferred to smaller models through distillation. Some studies explore the potential of leveraging LLMs to perform table-based reasoning. However, there has been no prior work focusing on table reasoning skills in smaller models specifically tailored for scientific table-to-text generation tasks. In this paper, we propose a novel table-based reasoning distillation approach, with the aim of distilling LLMs into tailored smaller models. Our experimental results have shown that a 220 million parameter model (Flan-T5-base) fine-tuned using distilled data, not only achieves a significant improvement compared to traditionally fine-tuned baselines, but also surpasses specific LLMs on a scientific table-to-text generation dataset. Our code is available at https://github.com/Bernard-Yang/DistillTableCoT.