import sys
import torch
import torch_musa
import torchvision
import torchvision.transforms as transforms
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim

## 2. построение сети
class Net(nn.Module):
    def __init__(self):
        super().__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 = torch.flatten(x, 1) # развернуть все измерения, кроме батча
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        x = self.fc3(x)
        return x

if __name__ == "__main__":
    ## 1. подготовка данных
    transform = transforms.Compose(
        [transforms.ToTensor(),
         transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])

    batch_size = 5
    trainset = torchvision.datasets.CIFAR10(root='./data', train=True,
                                            download=True, transform=transform)
    trainloader = torch.utils.data.DataLoader(trainset, batch_size=batch_size,
                                              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=batch_size,
                                             shuffle=False, num_workers=2)
    classes = ('plane', 'car', 'bird', 'cat',
               'deer', 'dog', 'frog', 'horse', 'ship', 'truck')
    device = torch.device("musa")

    net = Net().to(device)

    ## 3. определение функции потерь и оптимизатора
    criterion = nn.CrossEntropyLoss()
    optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)

    ## 4. обучение
    for epoch in range(2):  # проход по набору данных несколько раз
        running_loss = 0.0
        size = 0
        for i, data in enumerate(trainloader, 0):
            inputs, labels = data
            size += sys.getsizeof(inputs)
            inputs = inputs.to(device)
            outputs = net(inputs)
            loss = criterion(outputs, labels.to(device))
            loss.backward()
            optimizer.step()

            running_loss += loss.item()
            
            if i % 200 == 199 :
                print("размер данных ", size)
            
            if i % 2000 == 1999:    # вывод каждые 2000 мини-батчей
                print(f'[{epoch + 1}, {i + 1:5d}] loss: {running_loss / 2000:.3f}')
                running_loss = 0.0

    print('Обучение завершено')

    PATH = './cifar_net.pth'
    torch.save(net.state_dict(), PATH)

    net.load_state_dict(torch.load(PATH))

    ## 5. тестирование
    correct = 0
    total = 0
    with torch.no_grad():
        for data in testloader:
            images, labels = data
            outputs = net(images.to(device))
            _, predicted = torch.max(outputs.data, 1)
            total += labels.size(0)
            correct += (predicted == labels.to(device)).sum().item()

    print(f'Точность сети на 10000 тестовых изображениях: {100 * correct // total} %')