confopt: A Library for Implementation and Evaluation of Gradient-based One-Shot NAS Methods
This work addresses reproducibility and fair comparison issues for researchers in automated machine learning, though it is incremental as it builds on existing NAS methods.
The authors tackled the challenges of fragmented implementations and overreliance on the DARTS benchmark in gradient-based one-shot neural architecture search (NAS) by introducing confopt, an extensible library that streamlines development and evaluation, revealing a critical flaw in current assessment methods through new benchmarks and an evaluation protocol.
Gradient-based one-shot neural architecture search (NAS) has significantly reduced the cost of exploring architectural spaces with discrete design choices, such as selecting operations within a model. However, the field faces two major challenges. First, evaluations of gradient-based NAS methods heavily rely on the DARTS benchmark, despite the existence of other available benchmarks. This overreliance has led to saturation, with reported improvements often falling within the margin of noise. Second, implementations of gradient-based one-shot NAS methods are fragmented across disparate repositories, complicating fair and reproducible comparisons and further development. In this paper, we introduce Configurable Optimizer (confopt), an extensible library designed to streamline the development and evaluation of gradient-based one-shot NAS methods. Confopt provides a minimal API that makes it easy for users to integrate new search spaces, while also supporting the decomposition of NAS optimizers into their core components. We use this framework to create a suite of new DARTS-based benchmarks, and combine them with a novel evaluation protocol to reveal a critical flaw in how gradient-based one-shot NAS methods are currently assessed. The code can be found at https://github.com/automl/ConfigurableOptimizer.