English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Neural architecture search (NAS) methods have been proposed to release humanexperts from tedious architecture engineering. However, most current methodsare constrained in small-scale search due to the issue of computationalresources. Meanwhile, directly applying architectures searched on smalldatasets to large datasets often bears no performance guarantee. Thislimitation impedes the wide use of NAS on large-scale tasks. To overcome thisobstacle, we propose an elastic architecture transfer mechanism foraccelerating large-scale neural architecture search (EAT-NAS). In ourimplementations, architectures are first searched on a small dataset, e.g.,CIFAR-10. The best one is chosen as the basic architecture. The search processon the large dataset, e.g., ImageNet, is initialized with the basicarchitecture as the seed. The large-scale search process is accelerated withthe help of the basic architecture. What we propose is not only a NAS methodbut a mechanism for architecture-level transfer.  In our experiments, we obtain two final models EATNet-A and EATNet-B thatachieve competitive accuracies, 74.7% and 74.2% on ImageNet, respectively,which also surpass the models searched from scratch on ImageNet under the samesettings. For the computational cost, EAT-NAS takes only less than 5 days on 8TITAN X GPUs, which is significantly less than the computational consumption ofthe state-of-the-art large-scale NAS methods.

EAT-NAS: Elastic Architecture Transfer for Accelerating Large-scale Neural Architecture Search