Visformer学习

Vit

---

!pip install vit_pytorch

import torchvision
from torch.utils.tensorboard import SummaryWriter
from torch.utils.data import DataLoader
from torch import nn
from torch.utils.data import DataLoader

dataset_transform = torchvision.transforms.Compose([
    torchvision.transforms.Resize((32, 32)),
    torchvision.transforms.ToTensor(),
    torchvision.transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)),
])

train_data = torchvision.datasets.CIFAR10(root="../data", train=True, transform=dataset_transform,download=True)
                                          
test_data = torchvision.datasets.CIFAR10(root="../data", train=False, transform=dataset_transform,download=True)

train_data_size = len(train_data)
test_data_size = len(test_data)
print(train_data_size)
print(test_data_size)

print((train_data[1][0]).shape)

from torch.utils.data import DataLoader
train_dataloader=DataLoader(train_data,batch_size=64)
test_dataloader=DataLoader(train_data,batch_size=64)

for data in train_dataloader:
    imgs, targets = data
    print(imgs.shape)
    break



import torch
from vit_pytorch import ViT
# 创建ViT模型实例
v = ViT(
    image_size = 32,#原始图像尺寸
    patch_size = 8,#图像块尺寸
    num_classes = 10,#类别
    dim = 1024,#transformer隐变量维度大小
    depth = 6,#transformer编码器层
    heads = 16,#MSA头
    mlp_dim = 100,#多层感知机mlp维度
    dropout = 0.1,#失活比例
    emb_dropout = 0.1#嵌入层失活比例
)




if torch.cuda.is_available():
    v = v.cuda()

# 损失函数
loss_fn = nn.CrossEntropyLoss()
if torch.cuda.is_available():
    loss_fn = loss_fn.cuda()
# 优化器
# learning_rate = 0.01
# 1e-2=1 x (10)^(-2) = 1 /100 = 0.01
learning_rate = 1e-2
optimizer = torch.optim.SGD(v.parameters(), lr=learning_rate)

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

# 添加tensorboard
writer = SummaryWriter("../logs_train")

for i in range(epoch):
    print("-------第 {} 轮训练开始-------".format(i+1))

    # 训练步骤开始
    v.train()
    for data in train_dataloader:
        imgs, targets = data
        if torch.cuda.is_available():
            imgs = imgs.cuda()
            targets = targets.cuda()
        outputs = v(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
        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
    v.eval()
    total_test_loss = 0
    total_accuracy = 0
    with torch.no_grad():
        for data in test_dataloader:
            imgs, targets = data
            if torch.cuda.is_available():
                imgs = imgs.cuda()
                targets = targets.cuda()
            outputs = v(imgs)
            loss = loss_fn(outputs, targets)
            total_test_loss = total_test_loss + loss.item()
            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

    torch.save(tudui, "vit_{}.pth".format(i))
    print("模型已保存")

writer.close()

感慨就是训练的很慢，继续学习中