import torch
import torch.nn as nn
import torch.optim as optim
import numpy as np
import torchvision
from torchvision import datasets, models, transforms
import matplotlib.pyplot as plt
import torchvision.models as models
import matplotlib.pyplot as plt
import numpy as np
import copy
import time


def train_model(model, criterion, optimizer, scheduler, dataloader, dataset_size, num_epochs=25):
    since = time.time()

    best_model_wts = copy.deepcopy(model.state_dict())
    best_acc = 0.0

    for epoch in range(num_epochs):
        print('Epoch {}/{}'.format(epoch, num_epochs - 1))
        print('-' * 10)

        # Each epoch has a training and validation phase
        for phase in ['train', 'val']:
            if phase == 'train':
                model.train()  # Set model to training mode
            else:
                model.eval()   # Set model to evaluate mode

            running_loss = 0.0
            running_corrects = 0

            # Iterate over data.
            for inputs, labels in dataloader:

                # zero the parameter gradients
                optimizer.zero_grad()

                # forward
                # track history if only in train
                with torch.set_grad_enabled(phase == 'train'):
                    outputs = model(inputs)
                    _, preds = torch.max(outputs, 1)
                    loss = criterion(outputs, labels)

                    # backward + optimize only if in training phase
                    if phase == 'train':
                        loss.backward()
                        optimizer.step()

                # statistics
                running_loss += loss.item() * inputs.size(0)
                running_corrects += torch.sum(preds == labels.data)
            if phase == 'train':
                scheduler.step()

            epoch_loss = running_loss / dataset_size
            epoch_acc = running_corrects.double() / dataset_size

            print('{} Loss: {:.4f} Acc: {:.4f}'.format(
                phase, epoch_loss, epoch_acc))

            # deep copy the model
            if phase == 'val' and epoch_acc > best_acc:
                best_acc = epoch_acc
                best_model_wts = copy.deepcopy(model.state_dict())

        print()

    time_elapsed = time.time() - since
    print('Training complete in {:.0f}m {:.0f}s'.format(
        time_elapsed // 60, time_elapsed % 60))
    print('Best val Acc: {:4f}'.format(best_acc))

    # load best model weights
    model.load_state_dict(best_model_wts)
    return model


resnet18 = models.resnet18(pretrained=True)
resnet18.fc = nn.Linear(512, 10)
resnet18.conv1 = nn.Conv2d(1,64, kernel_size=(7,7), stride = (2,2), padding=(3,3), bias=False)
print(resnet18)
transform = transforms.Compose(
    [transforms.Resize((3,64)),
    transforms.ToTensor()])

mnist_testset = datasets.MNIST(root='./data', train=True, download=True, transform=transform)
# print((mnist_testset[0]))
testloader = torch.utils.data.DataLoader(mnist_testset, batch_size=4,shuffle=False, num_workers=2)

dataiter = iter(testloader)
# images, labels = dataiter.next()
# outputs = resnet18(images)
# _, predicted = torch.max(outputs, 1)
corr = 0
for ii, (inputs, labels) in enumerate(testloader):
    outputs = resnet18(inputs)
    _, predicted = torch.max(outputs, 1)
    x = predicted.cpu().detach().numpy()
    y = labels.cpu().detach().numpy()
    for i in range(4):
        if x[i] == y[i]:
            corr += 1
print(corr)