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 2007.07869cs.LGcs.CVstat.ML
收藏

长时间范围内基于约束的超参数优化

Gradient-based Hyperparameter Optimization Over Long Horizons

University of Edinburgh(爱丁堡大学)

Paul Micaelli, Amos Storkey

摘要:

基于一致性的超参数优化在Few-Shot元学习的上下文中获得了广泛的普及,但由于内存缩放和梯度退化问题,对于具有长视野(许多梯度步骤)的任务仍然是广泛不切实际的。一个常见的解决方法是在线学习超参数,但这会引入贪婪,从而导致性能显著下降。我们提出了前向模式差分与共享(FDS),一个简单而有效的算法,解决内存扩展问题与前向模式差分,和梯度退化问题,通过共享超参数,在时间上是连续的。我们提供了理论保证我们的算法的降噪性能,并证明其效率的经验,通过$\sim 10^4$梯度步骤展开优化区分。我们考虑在CIFAR-10上大的超参数搜索范围,在那里我们显著优于基于贪婪梯度的替代方案,同时与最先进的黑盒方法相比,实现了$\times 20$加速。代码可在:\url{https://github.com/polo5/FDS}

英文摘要:

Gradient-based hyperparameter optimization has earned a widespread popularity in the context of few-shot meta-learning, but remains broadly impractical for tasks with long horizons (many gradient steps), due to memory scaling and gradient degradation issues. A common workaround is to learn hyperparameters online, but this introduces greediness which comes with a significant performance drop. We propose forward-mode differentiation with sharing (FDS), a simple and efficient algorithm which tackles memory scaling issues with forward-mode differentiation, and gradient degradation issues by sharing hyperparameters that are contiguous in time. We provide theoretical guarantees about the noise reduction properties of our algorithm, and demonstrate its efficiency empirically by differentiating through $\sim 10^4$ gradient steps of unrolled optimization. We consider large hyperparameter search ranges on CIFAR-10 where we significantly outperform greedy gradient-based alternatives, while achieving $\times 20$ speedups compared to the state-of-the-art black-box methods. Code is available at: \url{https://github.com/polo5/FDS}

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。

↑