Generating Synthetic Datasets for Few-shot Prompt Tuning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Guo, Xu, Du, Zilin, Li, Boyang, Miao, Chunyan
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917802404741120
author Guo, Xu
Du, Zilin
Li, Boyang
Miao, Chunyan
author_facet Guo, Xu
Du, Zilin
Li, Boyang
Miao, Chunyan
contents A major limitation of prompt tuning is its dependence on large labeled training datasets. Under few-shot learning settings, prompt tuning lags far behind full-model fine-tuning, limiting its scope of application. In this paper, we leverage the powerful LLMs to synthesize task-specific labeled data for training the soft prompts. We first introduce a distribution-aligned weighted generator tuning (DawGen) method to encourage generating in-distribution data that aligns with the few-shot real data. Then, we train soft prompts on both synthetic and real datasets using a gradient surgery approach, which eliminates the conflicting gradients from different data sources. Experiments on seven sentence-pair classification datasets demonstrate the effectiveness of our proposed method for boosting prompt tuning in few-shot learning settings. Results on QQP, MRPC, and SICK datasets are even comparable to the performance of transfer learning from large real-world datasets, showing the promise of synthetic data as an alternative for enhancing soft prompt tuning.
format Preprint
id arxiv_https___arxiv_org_abs_2410_10865
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Generating Synthetic Datasets for Few-shot Prompt Tuning
Guo, Xu
Du, Zilin
Li, Boyang
Miao, Chunyan
Computation and Language
Artificial Intelligence
A major limitation of prompt tuning is its dependence on large labeled training datasets. Under few-shot learning settings, prompt tuning lags far behind full-model fine-tuning, limiting its scope of application. In this paper, we leverage the powerful LLMs to synthesize task-specific labeled data for training the soft prompts. We first introduce a distribution-aligned weighted generator tuning (DawGen) method to encourage generating in-distribution data that aligns with the few-shot real data. Then, we train soft prompts on both synthetic and real datasets using a gradient surgery approach, which eliminates the conflicting gradients from different data sources. Experiments on seven sentence-pair classification datasets demonstrate the effectiveness of our proposed method for boosting prompt tuning in few-shot learning settings. Results on QQP, MRPC, and SICK datasets are even comparable to the performance of transfer learning from large real-world datasets, showing the promise of synthetic data as an alternative for enhancing soft prompt tuning.
title Generating Synthetic Datasets for Few-shot Prompt Tuning
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2410.10865