PAODING: A High-fidelity Data-free Pruning Toolkit for Debloating Pre-trained Neural Networks
This work addresses the need for efficient model compression in machine learning, though it appears incremental as it builds on existing pruning techniques.
The authors tackled the problem of debloating pre-trained neural networks by introducing PAODING, a data-free pruning toolkit that iteratively removes neurons with minimal impact on output fidelity, resulting in significant model size reduction while preserving test accuracy and adversarial robustness.
We present PAODING, a toolkit to debloat pretrained neural network models through the lens of data-free pruning. To preserve the model fidelity, PAODING adopts an iterative process, which dynamically measures the effect of deleting a neuron to identify candidates that have the least impact to the output layer. Our evaluation shows that PAODING can significantly reduce the model size, generalize on different datasets and models, and meanwhile preserve the model fidelity in terms of test accuracy and adversarial robustness. PAODING is publicly available on PyPI via https://pypi.org/project/paoding-dl.