https://github.com/mitmul/chainer-cifar10

Various CNN models for CIFAR10 with Chainer
https://github.com/mitmul/chainer-cifar10

chainer cifar10 convnet convolutional-neural-networks deep-convolutional-networks deep-learning densenet network-in-network neural-networks residual-networks resnet vgg wide-residual-networks

Last synced: about 2 months ago
JSON representation

Various CNN models for CIFAR10 with Chainer

• Host: GitHub
• URL: https://github.com/mitmul/chainer-cifar10
• Owner: mitmul
• Created: 2015-06-09T14:39:43.000Z (over 9 years ago)
• Default Branch: master
• Last Pushed: 2018-12-06T12:35:09.000Z (almost 6 years ago)
• Last Synced: 2024-07-08T16:39:57.597Z (2 months ago)
• Topics: chainer, cifar10, convnet, convolutional-neural-networks, deep-convolutional-networks, deep-learning, densenet, network-in-network, neural-networks, residual-networks, resnet, vgg, wide-residual-networks
• Language: Python
• Homepage: http://chainer.org
• Size: 1.05 MB
• Stars: 138
• Watchers: 8
• Forks: 35
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Train various models on CIFAR10 with Chainer

## Requirements

- Python 3.5.1+ (not tested with Python2)

- pip packages:

    - chainer>=3.1.0

    - chainercv>=0.8.0

    - numpy>=1.10.1

    - matplotlib>=2.0.0

    - scikit-image>=0.13.1

    - opencv-python>=3.3.0

    - tabulate>=0.8.2

## Quick Start

```bash

MPLBACKEND=Agg python train.py

```

With full arguments:

```bash

MPLBACKEND=Agg python train.py \

--model_file models/wide_resnet.py \

--model_name WideResNet \

--batchsize 128 \

--training_epoch 500 \

--initial_lr 0.05 \

--lr_decay_rate 0.5 \

--lr_decay_epoch 70 \

--weight_decay 0.0005 \

--random_angle 15.0 \

--pca_sigma 25.5 \

--expand_ratio 1.2 \

--crop_size 28 28 \

--seed 0 \

--gpus 0 

```

## About data augmentation

It performs various data augmentation using [ChainerCV](https://github.com/chainer/chainercv). Provided operations are:

- Random rotating (using OpenCV or scikit-image)

- Random lighting

- Random LR-flipping

- Random zomming (a.k.a. expansion)

- Random cropping

See the details at `transform` function in `train.py`.

## Exprimental Results

| model_name   |   val/main/accuracy |   epoch |   batchsize | crop_size   |   expand_ratio |   pca_sigma |   random_angle |   weight_decay |   initial_lr |   lr_decay_rate |   lr_decay_epoch |

|:-------------|--------------------:|--------:|------------:|:------------|---------------:|------------:|---------------:|---------------:|-------------:|----------------:|-----------------:|

| LeNet5      |            0.860166 |     500 |         128 | [28, 28]    |            1.2 |        25.5 |             15 |         0.0005 |         0.01 |             0.5 |               50 |

| NIN          |            0.879351 |     500 |         128 | [28, 28]    |            1.2 |        25.5 |             15 |         0.0005 |         0.01 |             0.5 |              100 |

| VGG          |            0.934237 |     500 |         128 | [28, 28]    |            1.2 |        25.5 |             15 |         0.0005 |         0.05 |             0.5 |               50 |

| ResNet50     |            0.950455 |     500 |         128 | [28, 28]    |            1.2 |        25.5 |             15 |         0.0005 |         0.05 |             0.5 |               50 |

| DenseNet     |            0.944818 |     500 |         128 | [28, 28]    |            1.2 |        25.5 |             15 |         0.0005 |         0.05 |             0.5 |               50 |

| WideResNet   |            0.962322 |     500 |         128 | [28, 28]    |            1.2 |        25.5 |             15 |         0.0005 |         0.05 |             0.5 |               70 |

![](compare.png)