Large Language Models (LLMs) have shown enhanced capabilities of solving
novel tasks by reasoning step-by-step known as Chain-of-Thought (CoT)
reasoning; how can we instill the same capability of reasoning step-by-step on
unseen tasks into LMs that possess less than <100B parameters? To address this
question, we first introduce the CoT Collection, a new instruction-tuning
dataset that augments 1.88 million CoT rationales across 1,060 tasks. We show
that continually fine-tuning Flan-T5 (3B & 11B) with the CoT Collection enables
the 3B & 11B LMs to perform CoT better on unseen tasks, leading to an
improvement in the average zero-shot accuracy on 27 datasets of the
BIG-Bench-Hard benchmark by +4.34% and +2.44%, respectively. Furthermore, we
show that instruction tuning with CoT allows LMs to possess stronger few-shot
learning capabilities, resulting in an improvement of +2.97% and +2.37% on 4
domain-specific tasks over Flan-T5 (3B & 11B), respectively. We make our CoT
Collection data and our trained models publicly available at
this https URL

通过不同程度的 fine-tuning，基于 Flan-T5 的大型语言模型学习了 Chain-of-Thought 推理并表现出更强的 few-shot learning 能力，使得在 27 个数据集上的平均零 - shot 准确率提高了 4.34％和 2.44％，并在 4 个特定领域的任务上获得了进一步的改进。