TCM-FTP: Fine-Tuning Large Language Models for Herbal Prescription Prediction

TL;DR

This paper introduces TCM-FTP, a fine-tuning approach for large language models using a new TCM dataset, significantly improving the accuracy of herbal prescription prediction and dosage estimation.

Contribution

The paper presents DigestDS, a new TCM dataset, and TCM-FTP, a fine-tuning method that enhances LLM performance for herbal prescription prediction.

Findings

Achieves an F1-score of 0.8031 in prescription prediction

Demonstrates low-rank adaptation improves efficiency

Outperforms previous methods significantly

Abstract

Traditional Chinese medicine (TCM) has relied on specific combinations of herbs in prescriptions to treat various symptoms and signs for thousands of years. Predicting TCM prescriptions poses a fascinating technical challenge with significant practical implications. However, this task faces limitations due to the scarcity of high-quality clinical datasets and the complex relationship between symptoms and herbs. To address these issues, we introduce \textit{DigestDS}, a novel dataset comprising practical medical records from experienced experts in digestive system diseases. We also propose a method, TCM-FTP (TCM Fine-Tuning Pre-trained), to leverage pre-trained large language models (LLMs) via supervised fine-tuning on \textit{DigestDS}. Additionally, we enhance computational efficiency using a low-rank adaptation technique. Moreover, TCM-FTP incorporates data augmentation by permuting herbs within prescriptions, exploiting their order-agnostic nature. Impressively, TCM-FTP achieves an F1-score of 0.8031, significantly outperforming previous methods. Furthermore, it demonstrates remarkable accuracy in dosage prediction, achieving a normalized mean square error of 0.0604. In contrast, LLMs without fine-tuning exhibit poor performance. Although LLMs have demonstrated wide-ranging capabilities, our work underscores the necessity of fine-tuning for TCM prescription prediction and presents an effective way to accomplish this.