NAS evaluation is frustratingly hard

• URL: http://arxiv.org/abs/1912.12522v3
• Date: Thu, 13 Feb 2020 22:10:12 GMT
• Title: NAS evaluation is frustratingly hard
• Authors: Antoine Yang, Pedro M. Esperan\c{c}a, Fabio M. Carlucci
• Abstract summary: Neural Architecture Search (NAS) is an exciting new field which promises to be as much as a game-changer as Convolutional Neural Networks were in 2012. Comparison between different methods is still very much an open issue. Our first contribution is a benchmark of $8$ NAS methods on $5$ datasets.
• Score: 1.7188280334580197
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Neural Architecture Search (NAS) is an exciting new field which promises to be as much as a game-changer as Convolutional Neural Networks were in 2012. Despite many great works leading to substantial improvements on a variety of tasks, comparison between different methods is still very much an open issue. While most algorithms are tested on the same datasets, there is no shared experimental protocol followed by all. As such, and due to the under-use of ablation studies, there is a lack of clarity regarding why certain methods are more effective than others. Our first contribution is a benchmark of $8$ NAS methods on $5$ datasets. To overcome the hurdle of comparing methods with different search spaces, we propose using a method's relative improvement over the randomly sampled average architecture, which effectively removes advantages arising from expertly engineered search spaces or training protocols. Surprisingly, we find that many NAS techniques struggle to significantly beat the average architecture baseline. We perform further experiments with the commonly used DARTS search space in order to understand the contribution of each component in the NAS pipeline. These experiments highlight that: (i) the use of tricks in the evaluation protocol has a predominant impact on the reported performance of architectures; (ii) the cell-based search space has a very narrow accuracy range, such that the seed has a considerable impact on architecture rankings; (iii) the hand-designed macro-structure (cells) is more important than the searched micro-structure (operations); and (iv) the depth-gap is a real phenomenon, evidenced by the change in rankings between $8$ and $20$ cell architectures. To conclude, we suggest best practices, that we hope will prove useful for the community and help mitigate current NAS pitfalls. The code used is available at https://github.com/antoyang/NAS-Benchmark.

Related papers

• Delta-NAS: Difference of Architecture Encoding for Predictor-based Evolutionary Neural Architecture Search [5.1331676121360985]
  We craft an algorithm with the capability to perform fine-grain NAS at a low cost. We propose projecting the problem to a lower dimensional space through predicting the difference in accuracy of a pair of similar networks.
  arXiv  Detail & Related papers  (2024-11-21T02:43:32Z)
• The devil is in discretization discrepancy. Robustifying Differentiable NAS with Single-Stage Searching Protocol [2.4300749758571905]
  gradient-based methods suffer from the discretization error, which can severely damage the process of obtaining the final architecture. We introduce a novel single-stage searching protocol, which is not reliant on decoding a continuous architecture. Our results demonstrate that this approach outperforms other DNAS methods by achieving 75.3% in the searching stage on the Cityscapes validation dataset.
  arXiv  Detail & Related papers  (2024-05-26T15:44:53Z)
• Are Neural Architecture Search Benchmarks Well Designed? A Deeper Look Into Operation Importance [5.065947993017157]
  We conduct an empirical analysis of the widely used NAS-Bench-101, NAS-Bench-201 and TransNAS-Bench-101 benchmarks. We found that only a subset of the operation pool is required to generate architectures close to the upper-bound of the performance range. We consistently found convolution layers to have the highest impact on the architecture's performance.
  arXiv  Detail & Related papers  (2023-03-29T18:03:28Z)
• Generalizing Few-Shot NAS with Gradient Matching [165.5690495295074]
  One-Shot methods train one supernet to approximate the performance of every architecture in the search space via weight-sharing. Few-Shot NAS reduces the level of weight-sharing by splitting the One-Shot supernet into multiple separated sub-supernets. It significantly outperforms its Few-Shot counterparts while surpassing previous comparable methods in terms of the accuracy of derived architectures.
  arXiv  Detail & Related papers  (2022-03-29T03:06:16Z)
• $\beta$-DARTS: Beta-Decay Regularization for Differentiable Architecture Search [85.84110365657455]
  We propose a simple-but-efficient regularization method, termed as Beta-Decay, to regularize the DARTS-based NAS searching process. Experimental results on NAS-Bench-201 show that our proposed method can help to stabilize the searching process and makes the searched network more transferable across different datasets.
  arXiv  Detail & Related papers  (2022-03-03T11:47:14Z)
• BaLeNAS: Differentiable Architecture Search via the Bayesian Learning Rule [95.56873042777316]
  Differentiable Architecture Search (DARTS) has received massive attention in recent years, mainly because it significantly reduces the computational cost. This paper formulates the neural architecture search as a distribution learning problem through relaxing the architecture weights into Gaussian distributions. We demonstrate how the differentiable NAS benefits from Bayesian principles, enhancing exploration and improving stability.
  arXiv  Detail & Related papers  (2021-11-25T18:13:42Z)
• Prioritized Architecture Sampling with Monto-Carlo Tree Search [54.72096546595955]
  One-shot neural architecture search (NAS) methods significantly reduce the search cost by considering the whole search space as one network. In this paper, we introduce a sampling strategy based on Monte Carlo tree search (MCTS) with the search space modeled as a Monte Carlo tree (MCT) For a fair comparison, we construct an open-source NAS benchmark of a macro search space evaluated on CIFAR-10, namely NAS-Bench-Macro.
  arXiv  Detail & Related papers  (2021-03-22T15:09:29Z)
• Contrastive Neural Architecture Search with Neural Architecture Comparators [46.45102111497492]
  One of the key steps in Neural Architecture Search (NAS) is to estimate the performance of candidate architectures. Existing methods either directly use the validation performance or learn a predictor to estimate the performance. We propose a novel Contrastive Neural Architecture Search (CTNAS) method which performs architecture search by taking the comparison results between architectures as the reward.
  arXiv  Detail & Related papers  (2021-03-08T11:24:07Z)
• DA-NAS: Data Adapted Pruning for Efficient Neural Architecture Search [76.9225014200746]
  Efficient search is a core issue in Neural Architecture Search (NAS) We present DA-NAS that can directly search the architecture for large-scale target tasks while allowing a large candidate set in a more efficient manner. It is 2x faster than previous methods while the accuracy is currently state-of-the-art, at 76.2% under small FLOPs constraint.
  arXiv  Detail & Related papers  (2020-03-27T17:55:21Z)
• NAS-Bench-201: Extending the Scope of Reproducible Neural Architecture Search [55.12928953187342]
  We propose an extension to NAS-Bench-101: NAS-Bench-201 with a different search space, results on multiple datasets, and more diagnostic information. NAS-Bench-201 has a fixed search space and provides a unified benchmark for almost any up-to-date NAS algorithms. We provide additional diagnostic information such as fine-grained loss and accuracy, which can give inspirations to new designs of NAS algorithms.
  arXiv  Detail & Related papers  (2020-01-02T05:28:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.