arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2004.03340cs.CLcs.AIcs.LG
收藏

用CALM评估在线持续学习

Evaluating Online Continual Learning with CALM

Naver Labs Europe(Naver欧洲实验室);Pompeu Fabra University(庞培法布拉大学);CIIRC CTU(捷克理工大学计算机科学研究所)

Germán Kruszewski, Ionut-Teodor Sorodoc, Tomas Mikolov

摘要:

在线持续学习(OCL)研究在连续数据流上的学习,而不需要多次观察任何单个示例,这种设置更接近人类和系统必须“在野外”学习的经验。然而,通常可用的基准远离这些现实世界的条件,因为它们显式地表示不同的任务,缺乏潜在的相似性结构或假设不同示例之间的时间独立性。在这里,我们提出了一个基于语言建模的OCL新基准,其中输入在不同语言和域之间交替,而没有任何明确的界限。此外,我们提出了新的指标来研究灾难性遗忘在这种情况下,并评估多个基线模型的基础上组成的专家。最后,我们介绍了一种简单的门控技术,它可以学习不同输入之间的潜在相似性,从而提高专家产品模型的性能。

英文摘要:

Online Continual Learning (OCL) studies learning over a continuous data stream without observing any single example more than once, a setting that is closer to the experience of humans and systems that must learn "on-the-wild". Yet, commonly available benchmarks are far from these real-world conditions, because they explicitly signal different tasks, lack latent similarity structure or assume temporal independence between different examples. Here, we propose a new benchmark for OCL based on language modelling in which input alternates between different languages and domains without any explicit delimitation. Additionally, we propose new metrics to study catastrophic forgetting in this setting and evaluate multiple baseline models based on compositions of experts. Finally, we introduce a simple gating technique that learns the latent similarities between different inputs, improving the performance of a Products of Experts model.

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。

↑