Automated Deep Learning Using Neural Network Intelligence: Develop and Design PyTorch and TensorFlow Models Using Python

© The Author(s), under exclusive license to APress Media, LLC, part of Springer Nature 2022

I. GridinAutomated Deep Learning Using Neural Network Intelligencehttps://doi.org/10.1007/978-1-4842-8149-9_6

6. Model Pruning

Ivan Gridin¹  

(1)

Vilnius, Lithuania

 

Deep learning models have reached significant success in many real-life problems. A lot of devices use neural networks to perform everyday tasks. However, complex neural networks are computationally expensive. And not all devices have GPU processors to run deep learning models. Therefore, it would be helpful to perform model compression methods to reduce the model size and accelerate model performance without losing accuracy significantly. One of the main model compression techniques is model pruning. ...