TaskWeb: Selecting Better Source Tasks for Multi-task NLP

Joongwon Kim, Akari Asai, Gabriel Ilharco, Hannaneh Hajishirzi

Submission Type: Regular Long Paper

Submission Track: Machine Learning for NLP

Keywords: NLP, machine learning, task transfer, task selection, multi-task training

TL;DR: We quantify task relationships via pairwise task transfer, propose a new method for selecting helpful source tasks and apply it to multi-task training and improve zero-shot performances.

Abstract: Recent work in NLP has shown promising results in training models on large amounts of tasks to achieve better generalization. However, it is not well-understood how tasks are related, and how helpful training tasks can be chosen for a new task. In this work, we investigate whether knowing task relationships via pairwise task transfer improves choosing one or more source tasks that help to learn a new target task. We provide TaskWeb, a large-scale benchmark of pairwise task transfers for 22 NLP tasks using three different model types, sizes, and adaptation methods, spanning about 25,000 experiments. Then, we design a new method TaskShop based on our analysis of TaskWeb. TaskShop uses TaskWeb to estimate the benefit of using a source task for learning a new target task, and to choose a subset of helpful training tasks for multi-task training. Our method improves overall rankings and top-k precision of source tasks by 10% and 38%, respectively. We also use TaskShop to build much smaller multi-task training sets that improve zero-shot performances across 11 different target tasks by at least 4.3%.

Submission Number: 1977