Controlling Sparsity of Neural Network with Weight Constrained on Unit Lp-Sphere.
For more information, please contact author by: williamli_pro@163.com
The corresponding paper is under review.
Operating systems: Windows or Linux
Dependencies: CUDA, CuDNN, Python (version >= 3.5) Pytorch (version >= 1.2) and corresponding torchvision
After installed all of the requirements, please run demo.py
- Procedures of LpSS
- Weight distribution of a neuron with respect of p
- Summarization of test accuracy
| Dataset | nw | Epochs | s=0.5 | s=0.8 | s=0.9 | |
|---|---|---|---|---|---|---|
| UCI machine learning | DNA | 936195 | 40 | 0.9533 | 0.9233 | 0.9755 |
| Mushroom | 90562 | 20 | 1.0000 | 0.9890 | 0.7172 | |
| Climate | 76738 | 20 | 0.9714 | 0.9643 | 0.9214 | |
| MNIST | 170298 | 40 | 0.9957 | 0.9953 | 0.9946 | |
| Fashion-MNIST | 267514 | 50 | 0.9443 | 0.9391 | 0.9329 | |
| CIFAR-10 | 1154074 | 250 | 0.9423 | 0.9383 | 0.9310 | |
| Tiny ImageNet | 4644120 | 250 | 0.6377 | 0.6295 | 0.6042 |
Apache-2.0