Neural Architecture Search as Sparse Supernet
arXiv:2007.16112
Abstract
This paper aims at enlarging the problem of Neural Architecture Search (NAS) from Single-Path and Multi-Path Search to automated Mixed-Path Search. In particular, we model the NAS problem as a sparse supernet using a new continuous architecture representation with a mixture of sparsity constraints. The sparse supernet enables us to automatically achieve sparsely-mixed paths upon a compact set of nodes. To optimize the proposed sparse supernet, we exploit a hierarchical accelerated proximal gradient algorithm within a bi-level optimization framework. Extensive experiments on Convolutional Neural Network and Recurrent Neural Network search demonstrate that the proposed method is capable of searching for compact, general and powerful neural architectures.
arXiv commentAccepted to AAAI 2021
References in corpus (15)
- MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- Neural Architecture Search with Reinforcement Learning
- DARTS: Differentiable Architecture Search
- The composite absolute penalties family for grouped and hierarchical variable selection
- Group Sparse Regularization for Deep Neural Networks
- Neural Architecture Optimization
- PC-DARTS: Partial Channel Connections for Memory-Efficient Architecture Search
- DARTS+: Improved Differentiable Architecture Search with Early Stopping
- Learning the Number of Neurons in Deep Networks
- Progressive Differentiable Architecture Search: Bridging the Depth Gap between Search and Evaluation
- FasterSeg: Searching for Faster Real-time Semantic Segmentation
- BayesNAS: A Bayesian Approach for Neural Architecture Search
- GOLD-NAS: Gradual, One-Level, Differentiable
- One-Shot Neural Architecture Search via Compressive Sensing
- MixPath: A Unified Approach for One-shot Neural Architecture Search