Muppet: Massive Multi-task Representations with Pre-Finetuning
Armen Aghajanyan, Anchit Gupta, Akshat Shrivastava, Xilun Chen, Luke Zettlemoyer, Sonal Gupta
Abstract
We propose pre-finetuning, an additional large-scale learning stage between language model pre-training and fine-tuning. Pre-finetuning is massively multi-task learning (around 50 datasets, over 4.8 million total labeled examples), and is designed to encourage learning of representations that generalize better to many different tasks. We show that pre-finetuning consistently improves performance for pretrained discriminators (e.g. RoBERTa) and generation models (e.g. BART) on a wide range of tasks (sentence prediction, commonsense reasoning, MRC, etc.), while also significantly improving sample efficiency during fine-tuning. We also show that large-scale multi-tasking is crucial; pre-finetuning can hurt performance when few tasks are used up until a critical point (usually above 15) after which performance improves linearly in the number of tasks.
BibTeX
@inproceedings{aghajanyan-etal-2021-muppet,
title = "{M}uppet: Massive Multi-task Representations with Pre-Finetuning",
author = "Aghajanyan, Armen and
Gupta, Anchit and
Shrivastava, Akshat and
Chen, Xilun and
Zettlemoyer, Luke and
Gupta, Sonal",
editor = "Moens, Marie-Francine and
Huang, Xuanjing and
Specia, Lucia and
Yih, Scott Wen-tau",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.468/",
doi = "10.18653/v1/2021.emnlp-main.468",
pages = "5799--5811"
}