import torch
from torch import nn
import torch.nn.functional as F
import torchvision
import torchvision.transforms as transforms
import matplotlib.pyplot as plt
import numpy as np
#%%
transform = transforms.Compose(
    [transforms.ToTensor(),transforms.Normalize((0.5,0.5,0.5),(0.5,0.5,0.5))])
train_set  = torchvision.datasets.CIFAR10(root='E:\pytest\CIFAR_classficaion\data',train=True,download=True,
                                          transform=transform)
train_loader = torch.utils.data.DataLoader(train_set,batch_size = 4,
                                    shuffle = True,num_workers = 2)
test_set = torchvision.datasets.CIFAR10(root='E:\pytest\CIFAR_classficaion\data',train=False,download=True,
                                        transform=transform)
test_loader = torch.utils.data.DataLoader(test_set,batch_size = 4,
                                   shuffle = False,num_workers = 2)
classes = ('plane', 'car', 'bird', 'cat',
           'deer', 'dog', 'frog', 'horse', 'ship', 'truck')
#%%
def image_show(img):
    img = img.numpy()
    plt.imshow(np.transpose(img,(1,2,0)))
data_iter = iter(train_loader)
image,label = data_iter.__next__()
image_show(torchvision.utils.make_grid(image))
#%%
class net(nn.Module):
    def __init__(self):
        super(net,self).__init__()
        self.conv1 = nn.Conv2d(3,6,5)
        self.conv2 = nn.Conv2d(6,16,5)
        self.pool = nn.MaxPool2d(2,2)
        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()
#%%
import torch.optim as optim
l = nn.CrossEntropyLoss()
optim = optim.SGD(net.parameters(),lr = 0.001,momentum=0.9)

#%%
for epoch in range (2):
    losses = []
    for i,data in enumerate(train_loader,0):
        inputs,labels = data
        optim.zero_grad()
        outputs = net(inputs)
        loss = l(outputs,labels)
        loss.backward()
        optim.step()
        if i % 100 == 0:
            losses.append(loss.item())
            print(loss.item())
            plt.plot(losses)
            plt.show
print("over")
#%%
correct = 0
total = 0
with torch.no_grad():
    for image,labels in test_loader:
        outputs = net(image)
        loss = l(outputs,labels)
        _, predicted = torch.max(outputs.data, 1)
        total += labels.size(0)
        correct += (predicted == labels).sum().item()

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