In [8]:

    

import torch
import torchvision
import torchvision.transforms as transforms
%matplotlib inline


    

In [2]:

    

transform = transforms.Compose(
    [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=4,
                                          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=4,
                                         shuffle=False, num_workers=2)

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


    

    




Downloading http://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz
Files already downloaded and verified

In [9]:

    

import matplotlib.pyplot as plt
import numpy as np

# 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)))


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

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


    

    




 bird  bird truck   car






    

In [10]:

    

from torch.autograd import Variable
import torch.nn as nn
import torch.nn.functional as F


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)

    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.relu(self.fc2(x))
        x = self.fc3(x)
        return x


net = Net()


    

In [11]:

    

import torch.optim as optim

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


    

In [17]:

    

for epoch in range(10):  # 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), Variable(labels)

        # 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 % 2000 == 1999:    # print every 2000 mini-batches
            print('[%d, %5d] loss: %.3f' %
                  (epoch + 1, i + 1, running_loss / 2000))
            running_loss = 0.0

print('Finished Training')


    

    




[1,  2000] loss: 1.204
[1,  4000] loss: 1.185
[1,  6000] loss: 1.170
[1,  8000] loss: 1.182
[1, 10000] loss: 1.161
[1, 12000] loss: 1.174
[2,  2000] loss: 1.081
[2,  4000] loss: 1.092
[2,  6000] loss: 1.083
[2,  8000] loss: 1.096
[2, 10000] loss: 1.090
[2, 12000] loss: 1.086
[3,  2000] loss: 1.000
[3,  4000] loss: 1.002
[3,  6000] loss: 1.034
[3,  8000] loss: 1.029
[3, 10000] loss: 1.030
[3, 12000] loss: 1.018
[4,  2000] loss: 0.907
[4,  4000] loss: 0.942
[4,  6000] loss: 0.959
[4,  8000] loss: 0.968
[4, 10000] loss: 0.991
[4, 12000] loss: 0.999
[5,  2000] loss: 0.892
[5,  4000] loss: 0.906
[5,  6000] loss: 0.925
[5,  8000] loss: 0.913
[5, 10000] loss: 0.931
[5, 12000] loss: 0.931
[6,  2000] loss: 0.840
[6,  4000] loss: 0.871
[6,  6000] loss: 0.877
[6,  8000] loss: 0.864
[6, 10000] loss: 0.893
[6, 12000] loss: 0.913
[7,  2000] loss: 0.808
[7,  4000] loss: 0.820
[7,  6000] loss: 0.849
[7,  8000] loss: 0.840
[7, 10000] loss: 0.870
[7, 12000] loss: 0.858
[8,  2000] loss: 0.741
[8,  4000] loss: 0.800
[8,  6000] loss: 0.815
[8,  8000] loss: 0.818
[8, 10000] loss: 0.835
[8, 12000] loss: 0.860
[9,  2000] loss: 0.715
[9,  4000] loss: 0.766
[9,  6000] loss: 0.769
[9,  8000] loss: 0.820
[9, 10000] loss: 0.807
[9, 12000] loss: 0.826
[10,  2000] loss: 0.690
[10,  4000] loss: 0.740
[10,  6000] loss: 0.754
[10,  8000] loss: 0.791
[10, 10000] loss: 0.778
[10, 12000] loss: 0.817
Finished Training

In [18]:

    

dataiter = iter(testloader)
images, labels = dataiter.next()

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


    

    




GroundTruth:    cat  ship  ship plane






    

In [19]:

    

outputs = net(Variable(images))


    

In [20]:

    

_, predicted = torch.max(outputs.data, 1)

print('Predicted: ', ' '.join('%5s' % classes[predicted[j][0]]
                              for j in range(4)))


    

    




Predicted:    cat  ship plane plane

In [21]:

    

correct = 0
total = 0
for data in testloader:
    images, labels = data
    outputs = net(Variable(images))
    _, 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))


    

    




Accuracy of the network on the 10000 test images: 62 %

In [ ]: