Table 6 Comparison of the number of reserved subnet blocks.
From: Counterclockwise block-by-block knowledge distillation for neural network compression
The number of subnet blocks that are reserved | CIFAR-10 | Tiny-imagenet-200 | ||
|---|---|---|---|---|
VGG-16 (88.5%) | Resnet-18 (85.0%) | VGG-16 (59.4%) | Resnet-18 (57.3%) | |
0 | 87.9% | 85.3% | 58.8% | 54.8% |
1 | 88.4% | 85.6% | 59.4% | 56.0% |
2 | 88.1% | 85.6% | 5.9% | 55.6% |
