Towards Zero-Label Language Learning

TL;DR

This paper introduces a zero-label learning framework in NLP that uses synthetic data generated via few-shot prompts with pretrained models, achieving competitive results without human annotations.

Contribution

The paper proposes Unsupervised Data Generation (UDG), a novel method for creating high-quality training data from pretrained models without human labels, enabling zero-label learning and effective data augmentation.

Findings

Achieves comparable or better results than models trained on human-labeled data.

Sets new state-of-the-art on the SuperGLUE benchmark when combined with labeled data.

Demonstrates the effectiveness of synthetic data in training task-specific NLP models.

Abstract

This paper explores zero-label learning in Natural Language Processing (NLP), whereby no human-annotated data is used anywhere during training and models are trained purely on synthetic data. At the core of our framework is a novel approach for better leveraging the powerful pretrained language models. Specifically, inspired by the recent success of few-shot inference on GPT-3, we present a training data creation procedure named Unsupervised Data Generation (UDG), which leverages few-shot prompts to synthesize high-quality training data without real human annotations. Our method enables zero-label learning as we train task-specific models solely on the synthetic data, yet we achieve better or comparable results from strong baseline models trained on human-labeled data. Furthermore, when mixed with labeled data, our approach serves as a highly effective data augmentation procedure, achieving new state-of-the-art results on the SuperGLUE benchmark.