"""
    GPU训练 2：设置 device，分别在 模型、损失函数、图像组、标签组 后加 .to(device)
        # device = torch.device("cuda")
        # gis = gis.to(device)
        # loss_fn = loss_fn.to(device)
        # imgs = imgs.to(device)
        # targets = targets.to(device)
"""
import torch
import torchvision
from torch import nn

from torch.nn import Conv2d, MaxPool2d, ReLU, Sigmoid, Linear, Flatten, Sequential
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
import time

# from model.model20 import *

# 定义训练de设备（经过测试gpu:3060比cpu:i711s快了3.4倍左右）
# device = torch.device("cpu")  # 时间=277.95171666145325
device = torch.device("cuda")  # 时间=82.56803464889526(Windows11) 时间=52.36684012413025(Ubuntu20.04)

# 准备数据
train_data = torchvision.datasets.CIFAR10(root="./visionData", train=True, transform=torchvision.transforms.ToTensor(),
                                          download=True)
test_data = torchvision.datasets.CIFAR10(root="./visionData", train=False, transform=torchvision.transforms.ToTensor(),
                                         download=True)

# length 长度
train_data_size = len(test_data)
test_data_size = len(test_data)

print("训练数据集的长度为：{}".format(train_data_size))
print("测试数据集的长度为：{}".format(test_data_size))

# 利用DataLoader 加载数据
train_dataloader = DataLoader(train_data, batch_size=64)
test_dataloader = DataLoader(test_data, batch_size=64)


# 搭建神经网络
class Gisleung(nn.Module):
    def __init__(self):
        super(Gisleung, self).__init__()
        self.model = nn.Sequential(
            nn.Conv2d(3, 32, 5, 1, 2),
            nn.MaxPool2d(2),
            nn.Conv2d(32, 32, 5, 1, 2),
            nn.MaxPool2d(2),
            nn.Conv2d(32, 64, 5, 1, 2),
            nn.MaxPool2d(2),
            nn.Flatten(),
            nn.Linear(64 * 4 * 4, 64),
            nn.Linear(64, 10)
        )

    def forward(self, x):
        x = self.model(x)
        return x


# 实例化模型
gis = Gisleung()
# if torch.cuda.is_available():
#     gis = gis.cuda()
gis = gis.to(device)

# 损失函数
loss_fn = nn.CrossEntropyLoss()
# if torch.cuda.is_available():
#     loss_fn = loss_fn.cuda()
loss_fn = loss_fn.to(device)

# 优化器
learning_rate = 0.01  # 学习率
optimizer = torch.optim.SGD(gis.parameters(), lr=learning_rate)

# 设置训练网络的一些参数
total_train_step = 0  # 记录训练的次数
total_test_step = 0  # 记录测试的次数
epoch = 10  # 训练的轮数

# 添加tensorboard
writer = SummaryWriter("logs/020")
start_time = time.time()
for i in range(epoch):
    print("------第 {} 轮训练开始------".format(i + 1))

    # 训练步骤开始
    gis.train()
    for data in train_dataloader:
        imgs, targets = data
        # imgs = imgs.cuda()
        # targets = targets.cuda()
        imgs = imgs.to(device)
        targets = targets.to(device)

        outputs = gis(imgs)
        loss = loss_fn(outputs, targets)

        # 优化器优化模型
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        total_train_step = total_train_step + 1
        # 100次打印一次
        if total_train_step % 100 == 0:
            print("训练次数： {}，Loss：{}".format(total_train_step, loss.item()))
            writer.add_scalar("train_loss", loss.item(), total_train_step)

    # 测试步骤开始
    gis.eval()
    total_test_loss = 0
    total_accuracy = 0  # 整体正确的个数
    with torch.no_grad():
        for data in test_dataloader:
            imgs, targets = data
            # imgs = imgs.cuda()
            # targets = targets.cuda()
            imgs = imgs.to(device)
            targets = targets.to(device)

            outputs = gis(imgs)
            loss = loss_fn(outputs, targets)
            total_test_loss = total_test_loss + loss.item()
            # 计算正确率
            accuracy = (outputs.argmax(1) == targets).sum()
            total_accuracy = total_accuracy + accuracy

    print("整体测试集上的Loss：{}".format(total_test_loss))
    print("整体测试集上的正确率：{}".format(total_accuracy / test_data_size))

    writer.add_scalar("test_loss", total_test_loss, total_test_step)
    writer.add_scalar("test_accuracy", total_accuracy / test_data_size, total_test_step)

    total_test_step = total_test_step + 1

    # 保存每一次的模型
    torch.save(gis, "./model_save/gis_{}.pth".format(i))
    print("模型: gis_{}.pth 已保存".format(i))
    end_time = time.time()
    print("总共耗时：{}".format(end_time - start_time))

writer.close()