arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2210.05643cs.LGcs.CL
收藏

语言模型微调的核方法视角

A Kernel-Based View of Language Model Fine-Tuning

Princeton University(普林斯顿大学)

Sadhika Malladi, Alexander Wettig, Dingli Yu, Danqi Chen, Sanjeev Arora

AI总结:

本文用 NTK 与 Tensor Programs 刻画预训练语言模型在 Adam 微调下的核动态,并经 14 个 NLP 任务验证,解释提示式掩码预测及参数高效子空间微调的成功。

摘要:

通过微调预训练语言模型(LM)来解决 NLP 任务已成为标准做法,尤其是在低数据场景下。然而,人们对其实证成功的理论理解仍然很少,例如,为什么在几十个训练样本上微调一个具有 10^8 个或更多参数的模型不会导致过拟合。我们研究 Neural Tangent Kernel(NTK,神经正切核)是否能够描述预训练 LM 的微调过程;NTK 最初是作为一种模型提出的,用于研究在适当随机初始化下无限宽网络的梯度下降动态。本研究受到 NTK 在计算机视觉任务中良好性能的启发(Wei 等,2022)。我们将 NTK 形式体系扩展到 Adam,并利用 Tensor Programs(Yang,2020)刻画 NTK 视角能够描述预训练语言模型微调更新的条件。在 14 个 NLP 任务上的大量实验验证了我们的理论,并表明通过提示将下游任务表述为掩码词预测问题,通常会在微调期间引发基于核的动态。最后,我们利用这种核视角,为基于子空间的参数高效微调方法的成功提出了解释。

英文摘要:

It has become standard to solve NLP tasks by fine-tuning pre-trained language models (LMs), especially in low-data settings. There is minimal theoretical understanding of empirical success, e.g., why fine-tuning a model with $10^8$ or more parameters on a couple dozen training points does not result in overfitting. We investigate whether the Neural Tangent Kernel (NTK) - which originated as a model to study the gradient descent dynamics of infinitely wide networks with suitable random initialization - describes fine-tuning of pre-trained LMs. This study was inspired by the decent performance of NTK for computer vision tasks (Wei et al., 2022). We extend the NTK formalism to Adam and use Tensor Programs (Yang, 2020) to characterize conditions under which the NTK lens may describe fine-tuning updates to pre-trained language models. Extensive experiments on 14 NLP tasks validate our theory and show that formulating the downstream task as a masked word prediction problem through prompting often induces kernel-based dynamics during fine-tuning. Finally, we use this kernel view to propose an explanation for the success of parameter-efficient subspace-based fine-tuning methods.

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。

↑