RIFF: Learning to Rephrase Inputs for Few-shot Fine-tuning of Language Models
This work addresses the challenge of data efficiency in fine-tuning for NLP practitioners, but it is incremental as it builds on existing parameter-efficient methods.
The paper tackles the problem of improving few-shot fine-tuning of language models by rephrasing input texts, showing that using paraphrases at train and test time enhances performance beyond parameter-efficient fine-tuning alone on six few-shot text classification datasets.
Pre-trained Language Models (PLMs) can be accurately fine-tuned for downstream text processing tasks. Recently, researchers have introduced several parameter-efficient fine-tuning methods that optimize input prompts or adjust a small number of model parameters (e.g LoRA). In this study, we explore the impact of altering the input text of the original task in conjunction with parameter-efficient fine-tuning methods. To most effectively rewrite the input text, we train a few-shot paraphrase model with a Maximum-Marginal Likelihood objective. Using six few-shot text classification datasets, we show that enriching data with paraphrases at train and test time enhances the performance beyond what can be achieved with parameter-efficient fine-tuning alone. The code used for our experiments can be found at https://github.com/SaeedNajafi/RIFF.