Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective
Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective
复制标题
DOI：
--
复制
发表时间：
2021-02
期刊：
影响因子：
--
通讯作者：
Wuyang Chen;Xinyu Gong;Zhangyang Wang
中科院分区：
文献类型：
--
作者：
Wuyang Chen;Xinyu Gong;Zhangyang Wang
神经架构搜索（NAS）已经得到了爆炸性的研究，以自动发现性能最好的神经网络。目前的工作需要大量的超网训练或密集的架构评估，从而遭受沉重的资源消耗，并经常招致搜索偏见，由于截断训练或近似。我们能在不进行任何训练的情况下选择最好的神经架构，并消除搜索成本的一大部分吗？我们提供了一个肯定的答案，提出了一个新的框架，称为训练免费神经架构搜索（TE-NAS）。TE-NAS通过分析神经正切核（NTK）的频谱和输入空间中线性区域的数量来对架构进行排名。两者都是由深度网络的最新理论进展推动的，并且可以在没有任何训练和任何标签的情况下进行计算。我们证明：（1）这两个度量意味着神经网络的可训练性和表达性;（2）它们与网络的测试准确性密切相关。进一步，我们设计了一个基于修剪的NAS机制，以实现更灵活和上级之间的可训练性和表现力在搜索过程中的权衡。在NAS-Bench-201和DARTS搜索空间中，TE-NAS在CIFAR-10和ImageNet上分别使用一个1080Ti完成高质量搜索，但仅需0.5和4个GPU小时。我们希望我们的工作能够激发更多的尝试，将深度网络的理论发现与真实的NAS应用中的实际影响联系起来。代码可在：https://github.com/VITA-Group/TENAS.
Neural Architecture Search (NAS) has been explosively studied to automate the discovery of top-performer neural networks. Current works require heavy training of supernet or intensive architecture evaluations, thus suffering from heavy resource consumption and often incurring search bias due to truncated training or approximations. Can we select the best neural architectures without involving any training and eliminate a drastic portion of the search cost? We provide an affirmative answer, by proposing a novel framework called training-free neural architecture search (TE-NAS). TE-NAS ranks architectures by analyzing the spectrum of the neural tangent kernel (NTK) and the number of linear regions in the input space. Both are motivated by recent theory advances in deep networks and can be computed without any training and any label. We show that: (1) these two measurements imply the trainability and expressivity of a neural network; (2) they strongly correlate with the network's test accuracy. Further on, we design a pruning-based NAS mechanism to achieve a more flexible and superior trade-off between the trainability and expressivity during the search. In NAS-Bench-201 and DARTS search spaces, TE-NAS completes high-quality search but only costs 0.5 and 4 GPU hours with one 1080Ti on CIFAR-10 and ImageNet, respectively. We hope our work inspires more attempts in bridging the theoretical findings of deep networks and practical impacts in real NAS applications. Code is available at: https://github.com/VITA-Group/TENAS.