A NAS Benchmark for Diverse Tasks

Neural architecture search (NAS) benchmarks and methods prioritize performance on well-studied tasks, e.g., image classification on CIFAR and ImageNet. To mitigate this bias, NAS-Bench-360 is a benchmark suite for evaluating state-of-the-art NAS methods on a diverse set of tasks. The selection spans different application domains, dataset sizes, problem dimensionalities, and learning objectives. Please contact Renbo Tu ( or Nicholas Roberts ( if you have questions or wish to contribute to the benchmark.




Please cite the following if you use NAS-Bench-360:

NAS-Bench-360: Benchmarking Neural Architecture Search on Diverse Tasks
Renbo Tu*, Nicholas Roberts*, Mikhail Khodak, Junhong Shen, Frederic Sala, Ameet Talwalkar
		doi = {10.48550/ARXIV.2110.05668},
		url = {},
		author = {Tu, Renbo and Roberts, Nicholas and Khodak, Mikhail and Shen, Junhong and Sala, Frederic and Talwalkar, Ameet},
		title = {NAS-Bench-360: Benchmarking Neural Architecture Search on Diverse Tasks},
		publisher = {arXiv},
		year = {2021},


NAS-Bench-360 is released under the MIT license.