HuCurl:人类诱导的课程发现
HuCurl: Human-induced Curriculum Discovery
AI总结:
提出HuCurl课程发现框架,利用标注熵和损失度量样本难度,在课程空间中自动发现有效课程,证明最优课程常为非单调且可跨数据集与模型规模迁移。
摘要:
我们引入了课程发现问题,并描述了一个课程学习框架,该框架能够基于关于样本难度的先验知识,在课程空间中发现有效的课程。使用标注熵和损失作为难度度量,我们表明:(i)对于给定模型和数据集,表现最佳的已发现课程通常是非单调的,这与现有文献中的单调课程相反;(ii)流行的由易到难或由难到易的过渡课程往往存在表现不佳的风险;(iii)为较小数据集和模型发现的课程分别在较大数据集和模型上表现良好。所提出的框架涵盖了一些现有课程学习方法,并能够在多个NLP任务上发现优于它们的课程。
英文摘要:
We introduce the problem of curriculum discovery and describe a curriculum learning framework capable of discovering effective curricula in a curriculum space based on prior knowledge about sample difficulty. Using annotation entropy and loss as measures of difficulty, we show that (i): the top-performing discovered curricula for a given model and dataset are often non-monotonic as opposed to monotonic curricula in existing literature, (ii): the prevailing easy-to-hard or hard-to-easy transition curricula are often at the risk of underperforming, and (iii): the curricula discovered for smaller datasets and models perform well on larger datasets and models respectively. The proposed framework encompasses some of the existing curriculum learning approaches and can discover curricula that outperform them across several NLP tasks.
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。