Training code of pre-activation ResNet on ImageNet. It follows the setup in fb.resnet.torch (except for the weight decay) and gets similar performance (with much fewer lines of code). Models can be downloaded here.
| Model | Top 5 Error | Top 1 Error |
|---|---|---|
| ResNet 18 | 10.67% | 29.50% |
| ResNet 34 | 8.66% | 26.45% |
| ResNet 50 | 7.13% | 24.12% |
| ResNet 101 | 6.54% | 22.89% |
To train, just run:
./imagenet-resnet.py --data /path/to/original/ILSVRC --gpu 0,1,2,3 -d 18The speed is 1860 samples/s on 4 TitanX Pascal, and 1160 samples/s on 4 old TitanX, if your data is fast enough. See the tutorial on how to speed up your data.
This script only converts and runs ImageNet-ResNet{50,101,152} Caffe models released by Kaiming.
Example usage:
# convert caffe model to npy format
python -m tensorpack.utils.loadcaffe PATH/TO/{ResNet-101-deploy.prototxt,ResNet-101-model.caffemodel} ResNet101.npy
# run on an image
./load-resnet.py --load ResNet-101.npy --input cat.jpg --depth 101The converted models are verified on ILSVRC12 validation set. The per-pixel mean used here is slightly different from the original.
| Model | Top 5 Error | Top 1 Error |
|---|---|---|
| ResNet 50 | 7.89% | 25.03% |
| ResNet 101 | 7.16% | 23.74% |
| ResNet 152 | 6.81% | 23.28% |
Reproduce pre-activation ResNet on CIFAR10.
The train error shown here is a moving average of the error rate of each batch in training. The validation error here is computed on test set.
Also see a DenseNet implementation of the paper Densely Connected Convolutional Networks.