Contrastive Learning For Prompt-based Few-shot Language Learners

Jian Yiren, Gao Chongyang, Vosoughi Soroush. Arxiv 2022

[Paper] [Code]    
BERT Few Shot Fine Tuning GPT Has Code In Context Learning Language Modeling Masked Language Model Model Architecture Pretraining Methods Prompting Tools Training Techniques

The impressive performance of GPT-3 using natural language prompts and in-context learning has inspired work on better fine-tuning of moderately-sized models under this paradigm. Following this line of work, we present a contrastive learning framework that clusters inputs from the same class for better generality of models trained with only limited examples. Specifically, we propose a supervised contrastive framework that clusters inputs from the same class under different augmented “views” and repel the ones from different classes. We create different “views” of an example by appending it with different language prompts and contextual demonstrations. Combining a contrastive loss with the standard masked language modeling (MLM) loss in prompt-based few-shot learners, the experimental results show that our method can improve over the state-of-the-art methods in a diverse set of 15 language tasks. Our framework makes minimal assumptions on the task or the base model, and can be applied to many recent methods with little modification. The code will be made available at: https://github.com/yiren-jian/LM-SupCon.

Similar Work