A script to evaluate the computational workload of popular CNN models winning ILSVRC.
To use :
python benchmarkModel.py <.prototxt> <.caffemodel>
Please be sure to download the models using the appropriate script
Model | AlexNet | GoogleNet | VGG16 | VGG19 | ResNet50 | ResNet101 | ResNet-152 |
---|---|---|---|---|---|---|---|
Top1 err | 42.9 % | 31.3 % | 28.1 % | 27.3 % | 24.7% | 23.6% % | 23.0% |
Top5 err | 19.80 % | 10.07 % | 9.90 % | 9.00 % | 7.8 % | 7.1 % | 6.7 % |
conv layers | 5 | 57 | 13 | 16 | 53 | 104 | 155 |
conv workload (MACs) | 666 M | 1.58 G | 15.3 G | 19.5 G | 3.86 G | 7.57 G | 11.3 G |
conv parameters | 2.33 M | 5.97 M | 14.7 M | 20 M | 23.5 M | 42.4 M | 58 M |
pool layers | 3 | 14 | 5 | 5 | 2 | 2 | 2 |
FC layers | 3 | 1 | 3 | 3 | 1 | 1 | 1 |
FC workload (MACs) | 58.6 M | 1.02 M | 124 M | 124 M | 2.05 M | 2.05 M | 2.05 M |
FC parametrs | 58.6 M | 1.02 M | 124 M | 124 M | 2.05 M | 2.05 M | 2.05 M |
Total workload (MACs) | 724 M | 1.58 G | 15.5 G | 19.6 G | 3.86 G | 7.57 G | 11.3 G |
Total parameters | 61 M | 6.99 M | 138 M | 144 M | 25.5 M | 44.4 M | 60 M |