Towards Meta-Pruning via Optimal Transport
This addresses the efficiency and accuracy challenges in neural network compression for practitioners, though it appears incremental as it builds on existing pruning paradigms with a novel procedural twist.
The paper tackles the problem of structural pruning in neural networks, which often causes accuracy loss and requires fine-tuning, by introducing Intra-Fusion, a method that uses model fusion and Optimal Transport to achieve substantial accuracy recovery without fine-tuning, as demonstrated on datasets like CIFAR-10, CIFAR-100, and ImageNet.
Structural pruning of neural networks conventionally relies on identifying and discarding less important neurons, a practice often resulting in significant accuracy loss that necessitates subsequent fine-tuning efforts. This paper introduces a novel approach named Intra-Fusion, challenging this prevailing pruning paradigm. Unlike existing methods that focus on designing meaningful neuron importance metrics, Intra-Fusion redefines the overlying pruning procedure. Through utilizing the concepts of model fusion and Optimal Transport, we leverage an agnostically given importance metric to arrive at a more effective sparse model representation. Notably, our approach achieves substantial accuracy recovery without the need for resource-intensive fine-tuning, making it an efficient and promising tool for neural network compression. Additionally, we explore how fusion can be added to the pruning process to significantly decrease the training time while maintaining competitive performance. We benchmark our results for various networks on commonly used datasets such as CIFAR-10, CIFAR-100, and ImageNet. More broadly, we hope that the proposed Intra-Fusion approach invigorates exploration into a fresh alternative to the predominant compression approaches. Our code is available here: https://github.com/alexandertheus/Intra-Fusion.