PyTorch code for "Paying More Attention to Attention: Improving the Performance of
Convolutional Neural Networks via Attention Transfer" https://arxiv.org/abs/1612.03928
Conference paper at ICLR2017: https://openreview.net/forum?id=Sks9_ajex
What's in this repo so far:
- Activation-based AT code for CIFAR-10 experiments
- Code for ImageNet experiments (ResNet-18-ResNet-34 student-teacher)
Coming:
- grad-based AT
- Scenes and CUB activation-based AT code
- Pretrained with activation-based AT ResNet-18
The code uses PyTorch https://pytorch.org. Note that the original experiments were done using torch-autograd, we have so far validated that CIFAR-10 experiments are exactly reproducible in PyTorch, and are in process of doing so for ImageNet (results are very slightly worse in PyTorch, due to hyperparameters).
bibtex:
@inproceedings{Zagoruyko2017AT,
author = {Sergey Zagoruyko and Nikos Komodakis},
title = {Paying More Attention to Attention: Improving the Performance of
Convolutional Neural Networks via Attention Transfer},
booktitle = {ICLR},
url = {https://arxiv.org/abs/1612.03928},
year = {2017}}
First install PyTorch, then install torchnet:
pip install git+https://github.com/pytorch/tnt.git@master
then install other Python packages:
pip install -r requirements.txt
This section describes how to get the results in the table 1 of the paper.
First, train teachers:
python cifar.py --save logs/resnet_40_1_teacher --depth 40 --width 1
python cifar.py --save logs/resnet_16_2_teacher --depth 16 --width 2
python cifar.py --save logs/resnet_40_2_teacher --depth 40 --width 2
To train with activation-based AT do:
python cifar.py --save logs/at_16_1_16_2 --teacher_id resnet_16_2_teacher --beta 1e+3
To train with KD:
python cifar.py --save logs/kd_16_1_16_2 --teacher_id resnet_16_2_teacher --alpha 0.9
We plan to add AT+KD with decaying beta
to get the best knowledge transfer results soon.
We provide ResNet-18 pretrained model with activation based AT:
Model | val error |
---|---|
ResNet-18 | 30.4, 10.8 |
ResNet-18-ResNet-34-AT | 29.3, 10.0 |
Download link: https://s3.amazonaws.com/modelzoo-networks/resnet-18-at-export.pth
Model definition: https://github.com/szagoruyko/functional-zoo/blob/master/resnet-18-at-export.ipynb
Convergence plot:
Download pretrained weights for ResNet-34 (see also functional-zoo for more information):
wget https://s3.amazonaws.com/pytorch/h5models/resnet-34-export.hkl
Prepare the data following fb.resnet.torch and run training (e.g. using 2 GPUs):
python imagenet.py --imagenetpath ~/ILSVRC2012 --depth 18 --width 1 \
--teacher_params resnet-34-export.hkl --gpu_id 0,1 --ngpu 2 \
--beta 1e+3