We propose three novel pruning techniques to improve the cost and results of inference-aware Differentiable Neural Architecture Search (DNAS). First, we introduce Prunode, a stochastic bi-path building block for DNAS, which can search over inner hidden dimensions with O(1) memory and compute complexity. Second, we present an algorithm for pruning blocks within a stochastic layer of the SuperNet during the search. Third, we describe a novel technique for pruning unnecessary stochastic layers during the search. The optimized models resulting from the search are called PruNet and establishes a new state-of-the-art Pareto frontier for NVIDIA V100 in terms of inference latency for ImageNet Top-1 image classification accuracy. PruNet as a backbone also outperforms GPUNet and EfficientNet on the COCO object detection task on inference latency relative to mean Average Precision (mAP).
@article{arxiv.2209.11785,
title = {Tiered Pruning for Efficient Differentialble Inference-Aware Neural Architecture Search},
author = {Sławomir Kierat and Mateusz Sieniawski and Denys Fridman and Chen-Han Yu and Szymon Migacz and Paweł Morkisz and Alex-Fit Florea},
journal= {arXiv preprint arXiv:2209.11785},
year = {2023}
}