Scaling Instruction-Finetuned Language Models

作者：Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Eric Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, Albert Webson, Shixiang Gu, Zhuyun Dai, Mirac Süzgün, Xinyun Chen, Aakanksha Chowdhery, Castro-Ros, Alex, Pellat, Marie, Robinson, Kevin, Valter, Dasha, Sharan Narang, Gaurav Mishra, Adams Yu, Vincent Zhao, Yanping Huang, Andrew M. Dai, Hongkun Yu, Slav Petrov, Ed H., Jeff Dean, Jacob Devlin, Adam Roberts, Denny Zhou, Quoc V. Le, Wei, Jason · 发表于：arXiv (Cornell University) · 年份：2022 · DOI：10.48550/arxiv.2210.11416 · 被引用次数：1198 · 研究领域：Topic Modeling、Natural Language Processing Techniques、Text Readability and Simplification

Finetuning language models on a collection of datasets phrased as instructions has been shown to improve model performance and generalization to unseen tasks. In this paper we explore instruction finetuning with a particular focus on (1) scaling the number of tasks, (2) scaling the model size, and (3) finetuning on chain-of-thought data. We find that instruction finetuning with the above aspects dramatically improves performance on a variety of model classes (PaLM, T5, U-PaLM), prompting setups (zero-shot, few-shot, CoT), and evaluation benchmarks (MMLU, BBH, TyDiQA, MGSM, open-ended generation). For instance, Flan-PaLM 540B instruction-finetuned on 1.8K tasks outperforms PALM 540B by a large margin (+9.4% on average). Flan-PaLM 540B achieves state-of-the-art performance on several benchmarks, such as 75.2% on five-shot MMLU. We also publicly release Flan-T5 checkpoints, which achieve strong few-shot performance even compared to much larger models, such as PaLM 62B. Overall, instruction finetuning is a general method for improving the performance and usability of pretrained language models.