Training Task Experts through Retrieval Based Distillation
This addresses the data scarcity issue for specialized tasks in NLP, offering a novel approach to improve model performance, though it is incremental relative to existing distillation methods.
The paper tackles the problem of creating deployable models for specialized tasks where high-quality task-specific data is scarce, by introducing Retrieval Based Distillation (ReBase), which retrieves data from online sources to enhance diversity and generates Chain-of-Thought reasoning, resulting in performance improvements of up to 7.8% on SQuAD, 1.37% on MNLI, and 1.94% on BigBench-Hard.
One of the most reliable ways to create deployable models for specialized tasks is to obtain an adequate amount of high-quality task-specific data. However, for specialized tasks, often such datasets do not exist. Existing methods address this by creating such data from large language models (LLMs) and then distilling such knowledge into smaller models. However, these methods are limited by the quality of the LLMs output, and tend to generate repetitive or incorrect data. In this work, we present Retrieval Based Distillation (ReBase), a method that first retrieves data from rich online sources and then transforms them into domain-specific data. This method greatly enhances data diversity. Moreover, ReBase generates Chain-of-Thought reasoning and distills the reasoning capacity of LLMs. We test our method on 4 benchmarks and results show that our method significantly improves performance by up to 7.8% on SQuAD, 1.37% on MNLI, and 1.94% on BigBench-Hard.