AMC: AutoML for Model Compression and Acceleration on Mobile Devices. Model compression is a critical technique to efficiently deploy neural network models on mobile devices which have limited computation resources and tight power budgets. Conventional model compression techniques rely on hand-crafted heuristics and rule-based policies that require domain experts to explore the large design space trading off among model size, speed, and accuracy, which is usually sub-optimal and time-consuming. In this paper, we propose AutoML for Model Compression (AMC) which leverage reinforcement learning to provide the model compression policy. This learning-based compression policy outperforms conventional rule-based compression policy by having higher compression ratio, better preserving the accuracy and freeing human labor. Under 4x FLOPs reduction, we achieved 2.7% better accuracy than the handcrafted model compression policy for VGG-16 on ImageNet. We applied this automated, push-the-button compression pipeline to MobileNet and achieved 1.81x speedup of measured inference latency on an Android phone and 1.43x speedup on the Titan XP GPU, with only 0.1% loss of ImageNet Top-1 accuracy.
Keywords for this software
References in zbMATH (referenced in 5 articles )
Showing results 1 to 5 of 5.
- Chen, Xiaoli; Duan, Jinqiao; Karniadakis, George Em: Learning and meta-learning of stochastic advection-diffusion-reaction systems from sparse measurements (2021)
- Škrlj, Blaž; Martinc, Matej; Lavrač, Nada; Pollak, Senja: autoBOT: evolving neuro-symbolic representations for explainable low resource text classification (2021)
- Cui, Chunfeng; Zhang, Kaiqi; Daulbaev, Talgat; Gusak, Julia; Oseledets, Ivan; Zhang, Zheng: Active subspace of neural networks: structural analysis and universal attacks (2020)
- Neta Zmora, Guy Jacob, Lev Zlotnik, Bar Elharar, Gal Novik: Neural Network Distiller: A Python Package For DNN Compression Research (2019) arXiv
- Xinyu Gong, Shiyu Chang, Yifan Jiang, Zhangyang Wang: AutoGAN: Neural Architecture Search for Generative Adversarial Networks (2019) arXiv