argman / wide-residual-networks

95.8% and 80% on CIFAR-10 and CIFAR-100

Home Page:http://arxiv.org/abs/1605.07146

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Code for Wide Residual Networks

This code was used for experiments with Wide Residual Networks (BMVC 2016) http://arxiv.org/abs/1605.07146 by Sergey Zagoruyko and Nikos Komodakis.

Deep residual networks were shown to be able to scale up to thousands of layers and still have improving performance. However, each fraction of a percent of improved accuracy costs nearly doubling the number of layers, and so training very deep residual networks has a problem of diminishing feature reuse, which makes these networks very slow to train.

To tackle these problems, in this work we conduct a detailed experimental study on the architecture of ResNet blocks, based on which we propose a novel architecture where we decrease depth and increase width of residual networks. We call the resulting network structures wide residual networks (WRNs) and show that these are far superior over their commonly used thin and very deep counterparts.

For example, we demonstrate that even a simple 16-layer-deep wide residual network outperforms in accuracy and efficiency all previous deep residual networks, including thousand-layer-deep networks. We further show that WRNs achieve incredibly good results (e.g., achieving new state-of-the-art results on CIFAR-10, CIFAR-100 and SVHN) and train several times faster than pre-activation ResNets.

Test error (%, flip/translation augmentation) on CIFAR:

Method CIFAR-10 CIFAR-100
pre-ResNet-164 5.46 24.33
pre-ResNet-1001 4.92 22.71
WRN-28-10 4.17 20.5
WRN-28-10-dropout 4.39 20.0

See http://arxiv.org/abs/1605.07146 for details.

Installation

The code depends on Torch http://torch.ch. Follow instructions here and run:

luarocks install torchnet
luarocks install optnet
luarocks install iterm

We recommend installing CUDNN v5 for speed. Alternatively you can run on CPU or on GPU with OpenCL (coming).

For visualizing training curves we used ipython notebook with pandas and bokeh and suggest using anaconda.

Usage

Dataset support

The code supports loading simple datasets in torch format. We provide the following:

To whiten CIFAR-10 and CIFAR-100 we used the following scripts https://github.com/lisa-lab/pylearn2/blob/master/pylearn2/scripts/datasets/make_cifar10_gcn_whitened.py and then converted to torch using https://gist.github.com/szagoruyko/ad2977e4b8dceb64c68ea07f6abf397b and npy to torch converter https://github.com/htwaijry/npy4th.

We are running ImageNet experiments and will update the paper and this repo soon.

Training

We provide several scripts for reproducing results in the paper. Below are several examples.

model=wide-resnet widen_factor=4 depth=40 ./scripts/train_cifar.sh

This will train WRN-40-4 on CIFAR-10 whitened (supposed to be in datasets folder). This network achieves about the same accuracy as ResNet-1001 and trains in 6 hours on a single Titan X. Log is saved to logs/wide-resnet_$RANDOM$RANDOM folder with json entries for each epoch and can be visualized with itorch/ipython later.

For reference we provide logs for this experiment and ipython notebook to visualize the results. After running it you should see these training curves:

viz

Another example:

model=wide-resnet widen_factor=10 depth=28 dropout=0.3 dataset=./datasets/cifar100_whitened.t7 ./scripts/train_cifar.sh

This network achieves 20.0% error on CIFAR-100 in about a day on a single Titan X.

Multi-GPU is supported with nGPU=n parameter.

Other models

Additional models in this repo:

Implementation details

The code evolved from https://github.com/szagoruyko/cifar.torch. To reduce memory usage we use @fmassa's optimize-net, which automatically shares output and gradient tensors between modules. This keeps memory usage below 4 Gb even for our best networks. Also, it can generate network graph plots as the one for WRN-16-2 below.

About

95.8% and 80% on CIFAR-10 and CIFAR-100

http://arxiv.org/abs/1605.07146


Languages

Language:Lua 81.9%Language:Jupyter Notebook 10.6%Language:Python 5.7%Language:Shell 1.8%