Post-training deep neural network pruning via layer-wise calibration
This enables efficient neural network deployment on commodity hardware like edge devices, though it is incremental as it builds on existing pruning methods.
The paper tackles post-training weight pruning for deep neural networks, achieving state-of-the-art results with a ~1.5% accuracy drop at 50% sparsity using synthetic data and ~1% drop at 65% sparsity with real data on ResNet50/ImageNet.
We present a post-training weight pruning method for deep neural networks that achieves accuracy levels tolerable for the production setting and that is sufficiently fast to be run on commodity hardware such as desktop CPUs or edge devices. We propose a data-free extension of the approach for computer vision models based on automatically-generated synthetic fractal images. We obtain state-of-the-art results for data-free neural network pruning, with ~1.5% top@1 accuracy drop for a ResNet50 on ImageNet at 50% sparsity rate. When using real data, we are able to get a ResNet50 model on ImageNet with 65% sparsity rate in 8-bit precision in a post-training setting with a ~1% top@1 accuracy drop. We release the code as a part of the OpenVINO(TM) Post-Training Optimization tool.