基于BERT模型與RLHF的大語言模型協(xié)同校對方法研究

打開文本圖片集

中圖分類號：TP391.1；TP183 文獻(xiàn)標(biāo)識碼：A 文章編號：2096-4706（2025）11-0038-06

Research on Collaborative Proofreading Method of Large Language Model Based on BERT Model and RLHF

WU Bian1, YANG Zhengtan2,LI Xiang (1.StateGrid Hubei Electric PowerCo.,Ltd.,Wuhan 430048,China; 2.Wuhan Optics Valley Information Technology Co.,Ltd.,Wuhan 430206, China)

Abstract: The auracy of document proofreading has always faces challenges at the level of complex logic.In order toaleviatethepresureonwritersandfront-linestaff,thisstudyproposesaproofreadingmethodbasedonmulti-model collaboration.Theword-by-wordlabelisgeneratedbyfine-tuningBERTmodel,andtheLargeLanguageModelisfine-tuning usingLoRA tocompensatefordeficienciesindeeperrorunderstanding.ThePPOalgorithm isused tooptimizethedecisionmaking processof te model to met the needsof different scenarios.The multi-modeloutputresultsare integrated through XGBoost toavoidundereporting and misreporting.The experimentalresultsshow thatthis methodhassignifcant advantages in improving the quality and accuracy of document proofreading.

Keywords: document proofreading; BERT;LLM; PPO; XGBoost

0 引言

公文作為黨政機(jī)關(guān)、企事業(yè)單位乃至學(xué)術(shù)機(jī)構(gòu)日常工作中的重要工具，承擔(dān)著信息傳遞、決策指導(dǎo)和政策執(zhí)行的關(guān)鍵任務(wù)[1。（剩余9708字）