In this paper, we present a novel sensitivity-based filter pruning algorithm (SbF-Pruner) to learn the importance scores of filters of each layer end-to-end. Our method learns the scores from the filter weights, enabling it to account for the correlations between the filters of each layer. Moreover, by training the pruning scores of all layers simultaneously our method can account for layer interdependencies, which is essential to find a performant sparse sub-network. Our proposed method can train and generate a pruned network from scratch in a straightforward, one-stage training process without requiring a pretrained network. Ultimately, we do not need layer-specific hyperparameters and pre-defined layer budgets, since SbF-Pruner can implicitly determine the appropriate number of channels in each layer. Our experimental results on different network architectures suggest that SbF-Pruner outperforms advanced pruning methods. Notably, on CIFAR-10, without requiring a pretrained baseline network, we obtain 1.02% and 1.19% accuracy gain on ResNet56 and ResNet110, compared to the baseline reported for state-of-the-art pruning algorithms. This is while SbF-Pruner reduces parameter-count by 52.3% (for ResNet56) and 54% (for ResNet101), which is better than the state-of-the-art pruning algorithms with a high margin of 9.5% and 6.6%.
翻译:在本文中,我们展示了一种新的基于敏感度的过滤过滤运行算法( SbF- Pruner ), 以学习每个层端到端的过滤器重要分数。 我们的方法从过滤器重量中学习分数, 使其能够计算每个层过滤器的关联性。 此外, 通过同时培训所有层的分数, 我们的方法可以考虑到层间相互依存性, 这对于找到一个性能稀疏的子网络至关重要。 我们建议的方法可以在不要求预先训练网络的情况下, 在一个直截了当的、 一阶段的培训过程中, 训练并产生一个纯净化的网络网络。 最后, 我们不需要针对具体层的超参数和预定义的层预算, 因为 SbF- Pruner可以暗中确定每个层过滤器之间的相关性。 我们在不同网络结构上的实验结果显示, SbF- Prunerger 超越了先进的运行方法。 值得注意的是, 我们不需要事先训练的基线网络, 我们可以在ResNet56 和 ResNet110 和 ResNet- 110 1010 上获得1.02% 的精准率收益收益, 而Sb- remart- real_ real prain_ real- prain% stranscurrus- strading pressional_ real_ real_ real_ redufal- s