toy_cifar.py

#!/usr/bin/python

import torch
import torchvision
import torchvision.transforms as transforms
from torch.autograd import Variable
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import matplotlib.pyplot as plt
import numpy as np

'''
PyTorch practice.

This is almost the same with 
http://pytorch.org/tutorials/beginner/blitz/cifar10_tutorial.html
'''

BATCHSIZE = 64

# functions to show an image
def imshow(img):
    img = img / 2 + 0.5     # unnormalize
    npimg = img.numpy()
    plt.imshow(np.transpose(npimg, (1, 2, 0)))


# Dataloader
transform = transforms.Compose(
    [transforms.RandomHorizontalFlip(),         # Random Flip (FIXME)
     transforms.ToTensor(),
     transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])

trainset = torchvision.datasets.CIFAR10(root='./data', train=True,
                                        download=True, transform=transform)
trainloader = torch.utils.data.DataLoader(trainset, batch_size=BATCHSIZE,
                                          shuffle=True, num_workers=2)

testset = torchvision.datasets.CIFAR10(root='./data', train=False,
                                       download=True, transform=transform)
testloader = torch.utils.data.DataLoader(testset, batch_size=BATCHSIZE,
                                         shuffle=False, num_workers=2)

classes = ('plane', 'car', 'bird', 'cat',
           'deer', 'dog', 'frog', 'horse', 'ship', 'truck')


# get some random training images
dataiter = iter(trainloader)
images, labels = dataiter.next()

# show images
imshow(torchvision.utils.make_grid(images))
plt.show()
# print labels
print(' '.join('%5s' % classes[labels[j]] for j in range(4)))


# Define network
class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(3, 6, 5)
        self.pool = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(6, 16, 5)
        self.fc1 = nn.Linear(16 * 5 * 5, 120)
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 10)

    # Added dropout
    def forward(self, x):
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        x = x.view(-1, 16 * 5 * 5)
        x = F.relu(self.fc1(x))
        x = F.dropout(x, training=self.training)
        x = F.relu(self.fc2(x))
        x = F.dropout(x, training=self.training)
        x = self.fc3(x)
        return x


# Instantiate the network
net = Net()
net.cuda()   # Enable GPU operations

criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)




# Training
for epoch in range(100):  # loop over the dataset multiple times

    running_loss = 0.0
    for i, data in enumerate(trainloader, 0):
        # get the inputs
        inputs, labels = data

        # wrap them in Variable
        inputs, labels = Variable(inputs.cuda()), Variable(labels.cuda())

        # zero the parameter gradients
        optimizer.zero_grad()

        # forward + backward + optimize
        outputs = net(inputs)
        loss = criterion(outputs, labels)
        loss.backward()
        optimizer.step()

        # print statistics
        running_loss += loss.data[0]
        if i % 100 == 99:    # print every 100 mini-batches
            print('[%d, %5d] loss: %.3f' %
                  (epoch + 1, i + 1, running_loss / 2000))
            running_loss = 0.0


    # Print test results after each epoch
    correct = 0
    total = 0
    for data in testloader:
        images, labels = data
        outputs = net(Variable(images.cuda())).cpu()
        _, predicted = torch.max(outputs.data, 1)
        total += labels.size(0)
        correct += (predicted == labels).sum()

    print('test_acc: %d %%' % (100 * correct / total))

print('Finished Training')


# Testing
dataiter = iter(testloader)
images, labels = dataiter.next()
net.eval()    # Test mode (dropout)

# print images
imshow(torchvision.utils.make_grid(images))
plt.show()
print('GroundTruth: ', ' '.join('%5s' % classes[labels[j]] for j in range(4)))


correct = 0
total = 0
for data in testloader:
    images, labels = data
    outputs = net(Variable(images.cuda())).cpu()
    _, predicted = torch.max(outputs.data, 1)
    total += labels.size(0)
    correct += (predicted == labels).sum()

print('Accuracy of the network on the 10000 test images: %d %%' % (
    100 * correct / total))


class_correct = list(0. for i in range(10))
class_total = list(0. for i in range(10))
for data in testloader:
    images, labels = data
    outputs = net(Variable(images.cuda())).cpu()
    _, predicted = torch.max(outputs.data, 1)
    c = (predicted == labels).squeeze()
    for i in range(4):
        label = labels[i]
        class_correct[label] += c[i]
        class_total[label] += 1


for i in range(10):
    print('Accuracy of %5s : %2d %%' % (
        classes[i], 100 * class_correct[i] / class_total[i]))