Meta Architecture Search

Abstract

Neural Architecture Search (NAS) has been quite successful in constructing state-of-the-art models on a variety of tasks. Unfortunately, the computational cost can make it difficult to scale. In this paper, we make the first attempt to study Meta Architecture Search which aims at learning a task-agnostic representation that can be used to speed up the process of architecture search on a large number of tasks. We propose the Bayesian Meta Architecture SEarch (BASE) framework which takes advantage of a Bayesian formulation of the architecture search problem to learn over an entire set of tasks simultaneously. We show that on Imagenet classification, we can find a model that achieves 25.7% top-1 error and 8.1% top-5 error by adapting the architecture in less than an hour from an 8 GPU days pretrained meta-network. By learning a good prior for NAS, our method dramatically decreases the required computation cost while achieving comparable performance to current state-of-the-art methods - even finding competitive models for unseen datasets with very quick adaptation. We believe our framework will open up new possibilities for efficient and massively scalable architecture search research across multiple tasks.

11 pages, 4 figures, 4 tables, 4 pages of appendix; NeurIPS 2019

Cited by in corpus (15)

• NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing
• Design Space for Graph Neural Networks
• Learning to Branch for Multi-Task Learning
• Weight-Sharing Neural Architecture Search: A Battle to Shrink the Optimization Gap
• Deep Transformers with Latent Depth
• Speedy Performance Estimation for Neural Architecture Search
• TF-NAS: Rethinking Three Search Freedoms of Latency-Constrained Differentiable Neural Architecture Search
• Learning Versatile Neural Architectures by Propagating Network Codes
• HourNAS: Extremely Fast Neural Architecture Search Through an Hourglass Lens
• Rapid Neural Architecture Search by Learning to Generate Graphs from Datasets
• Task-Adaptive Neural Network Search with Meta-Contrastive Learning
• TransNAS-Bench-101: Improving Transferability and Generalizability of Cross-Task Neural Architecture Search
• HR-NAS: Searching Efficient High-Resolution Neural Architectures with Lightweight Transformers
• Rapid Structural Pruning of Neural Networks with Set-based Task-Adaptive Meta-Pruning
• Differentiable Neural Architecture Search with Morphism-based Transformable Backbone Architectures