Meta Learning via Learned Loss-猫眼课题宝

DOI：

10.1109/icpr48806.2021.9412010

发表时间：

2019-06

期刊：

2020 25th International Conference on Pattern Recognition (ICPR)

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

Yevgen Chebotar;Artem Molchanov;Sarah Bechtle;L. Righetti;Franziska Meier;G. Sukhatme

通常，损失函数、正则化机制和训练参数模型的其他重要方面都是从有限的选项中选择的。在本文中，我们朝着自动化这个过程迈出了第一步，目的是生产更快，更强大的训练模型。具体来说，我们提出了一种元学习方法来学习参数损失函数，可以在不同的任务和模型架构中推广。我们开发了一个用于“元训练”这种损失函数的管道，旨在最大限度地提高在它们下训练的模型的性能。我们学习的损失产生的损失景观显着改善了监督和强化学习任务中的原始任务特定损失。此外，我们还证明了我们的元学习框架足够灵活，可以在元训练时纳入额外的信息。该信息形成学习损失函数，使得环境不需要在元测试时间期间提供该信息。我们在https://sites.google.com/view/mlthree上提供代码

Typically, loss functions, regularization mechanisms and other important aspects of training parametric models are chosen heuristically from a limited set of options. In this paper, we take the first step towards automating this process, with the view of producing models which train faster and more robustly. Concretely, we present a meta-learning method for learning parametric loss functions that can generalize across different tasks and model architectures. We develop a pipeline for “meta-training” such loss functions, targeted at maximizing the performance of the model trained under them. The loss landscape produced by our learned losses significantly improves upon the original task-specific losses in both supervised and reinforcement learning tasks. Furthermore, we show that our meta-learning framework is flexible enough to incorporate additional information at meta-train time. This information shapes the learned loss function such that the environment does not need to provide this information during meta-test time. We make our code available at https://sites.google.com/view/mlthree

Meta Learning via Learned Loss

文献摘要

被引文献

相似文献