从实例训练到指令学习：基于指令生成任务适配器
From Instance Training to Instruction Learning: Task Adapters Generation from Instructions
AI总结：
本文提出TAGI框架，依据任务指令以参数生成方式构建任务适配器，并通过知识蒸馏与两阶段超网络训练实现跨任务泛化，在多个基准上以更低计算成本匹敌或超越元学习及超网络基线。
摘要：
大型语言模型（LLM）已通过指令微调（IFT）获得了解决通用任务的能力。然而，IFT仍然严重依赖大量任务数据的实例训练，这极大限制了LLM在真实场景中的适应性；在这些场景中，带标注的任务实例稀缺，而更广泛的任务泛化变得至关重要。与LLM不同，人类获取技能并完成任务不仅依靠重复练习，还依靠理解并遵循指令性指南。本文致力于模拟人类学习以解决实例训练的不足，重点关注指令学习以增强跨任务泛化。在此背景下，我们提出了Task Adapters Generation from Instructions（TAGI，基于指令生成任务适配器），它根据给定任务指令，以参数生成方式自动构建特定任务模型，而无需针对未见任务重新训练。具体而言，我们利用知识蒸馏，通过对齐二者之间的标签、输出logits和适配器参数，增强通过Learning with Instruction开发的TAGI与通过Training with Instance开发的特定任务模型之间的一致性。TAGI通过包含超网络预训练和微调的两阶段训练过程获得跨任务泛化能力。我们在Super-Natural Instructions和P3数据集上评估了TAGI。实验结果表明，TAGI能够匹敌甚至超越传统元训练模型和其他超网络模型，同时显著降低计算需求。
英文摘要：
Large language models (LLMs) have acquired the ability to solve general tasks by utilizing instruction finetuning (IFT). However, IFT still relies heavily on instance training of extensive task data, which greatly limits the adaptability of LLMs to real-world scenarios where labeled task instances are scarce and broader task generalization becomes paramount. Contrary to LLMs, humans acquire skills and complete tasks not merely through repeated practice but also by understanding and following instructional guidelines. This paper is dedicated to simulating human learning to address the shortcomings of instance training, focusing on instruction learning to enhance cross-task generalization. Within this context, we introduce Task Adapters Generation from Instructions (TAGI), which automatically constructs the task-specific model in a parameter generation manner based on the given task instructions without retraining for unseen tasks. Specifically, we utilize knowledge distillation to enhance the consistency between TAGI developed through Learning with Instruction and task-specific models developed through Training with Instance, by aligning the labels, output logits, and adapter parameters between them. TAGI is endowed with cross-task generalization capabilities through a two-stage training process that includes hypernetwork pretraining and finetuning. We evaluate TAGI on the Super-Natural Instructions and P3 datasets. The experimental results demonstrate that TAGI can match or even outperform traditional meta-trained models and other hypernetwork models, while significantly reducing computational requirements.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
