BriefGPT.xyz
Apr, 2024
使用PEFT和合成数据增强低资源LLMs分类
Enhancing Low-Resource LLMs Classification with PEFT and Synthetic Data
HTML
PDF
Parth Patwa, Simone Filice, Zhiyu Chen, Giuseppe Castellucci, Oleg Rokhlenko...
TL;DR
提出了一种方法，使大型语言模型在0-shot文本分类任务中成为高效的文本分类器，并在低资源环境下获得了竞争性结果。
Abstract
large language models
(LLMs) operating in 0-shot or few-shot settings achieve competitive results in Text Classification tasks.
in-context learning
(ICL) typically achieves better accuracy than the 0-shot setting
→