# Multi-fidelity Neural Architecture Search with Knowledge Distillation

Source: [Multi-fidelity Neural Architecture Search with Knowledge Distillation](<https://www.sota2.com/research/paper/68fa69ef1e209131cea250d2>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T23:12:37.633Z

Authors: Ilya Trofimov, Nikita Klyuchnikov, Mikhail Salnikov, Alexander Filippov, Evgeny Burnaev

Published: 2020-06-15T12:32:38.000Z

## Abstract

Neural architecture search (NAS) targets at finding the optimal architecture of a neural network for a problem or a family of problems. Evaluations of neural architectures are very time-consuming. One of the possible ways to mitigate this issue is to use low-fidelity evaluations, namely training on a part of a dataset, fewer epochs, with fewer channels, etc. In this paper, we propose a bayesian multi-fidelity method for neural architecture search: MF-KD. The method relies on a new approach to low-fidelity evaluations of neural architectures by training for a few epochs using a knowledge distillation. Knowledge distillation adds to a loss function a term forcing a network to mimic some teacher network. We carry out experiments on CIFAR-10, CIFAR-100, and ImageNet-16-120. We show that training for a few epochs with such a modified loss function leads to a better selection of neural architectures than training for a few epochs with a logistic loss. The proposed method outperforms several state-of-the-art baselines.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2006.08341>)

[PDF](<https://arxiv.org/pdf/2006.08341.pdf>)
