ESPN: Extremely Sparse Pruned Networks
arXiv:2006.15741
Abstract
Deep neural networks are often highly overparameterized, prohibiting their use in compute-limited systems. However, a line of recent works has shown that the size of deep networks can be considerably reduced by identifying a subset of neuron indicators (or mask) that correspond to significant weights prior to training. We demonstrate that an simple iterative mask discovery method can achieve state-of-the-art compression of very deep networks. Our algorithm represents a hybrid approach between single shot network pruning methods (such as SNIP) with Lottery-Ticket type approaches. We validate our approach on several datasets and outperform several existing pruning approaches in both test accuracy and compression ratio.
References in corpus (21)
- Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations
- AMC: AutoML for Model Compression and Acceleration on Mobile Devices
- A Survey of Model Compression and Acceleration for Deep Neural Networks
- Compressing Neural Networks with the Hashing Trick
- Model compression via distillation and quantization
- Comparing Rewinding and Fine-tuning in Neural Network Pruning
- Deep Rewiring: Training very sparse deep networks
- Picking Winning Tickets Before Training by Preserving Gradient Flow
- Stabilizing the Lottery Ticket Hypothesis
- Memory Bounded Deep Convolutional Networks
- Faster gaze prediction with dense networks and Fisher pruning
- Compressing Neural Networks using the Variational Information Bottleneck
- Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization
- Deconstructing Lottery Tickets: Zeros, Signs, and the Supermask
- Deep Model Compression: Distilling Knowledge from Noisy Teachers
- CondenseNet: An Efficient DenseNet using Learned Group Convolutions
- Soft Weight-Sharing for Neural Network Compression
- Fixing the train-test resolution discrepancy: FixEfficientNet
- Winning the Lottery with Continuous Sparsification
- Movement Pruning: Adaptive Sparsity by Fine-Tuning
- Pruning via Iterative Ranking of Sensitivity Statistics