Blair

1 导入包

from torchvision import datasets,transforms
from torch.utils.data import DataLoader
import matplotlib.pyplot as plt
import torchvision
from torch import nn
import numpy as np
import torch

2 加载数据集

transformation =transforms.Compose([
    transforms.ToTensor()  #转换到Tensor，并且转换为0-1之间,将channel 放到第一个纬度
])
train_ds = datasets.MNIST('data/',train = True,transform = transformation,download = True)
test_ds = datasets.MNIST('data/',train = False,transform = transformation,download = True)
# len(train_ds)
# len(test_ds)

3 加载loader

train_loader = DataLoader(train_ds,batch_size =64 ,shuffle = True,num_workers = 16)
test_loader = DataLoader(test_ds,batch_size =256 ,shuffle = False,num_workers = 16)

4 创建模型

class Model(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(28*28,128)
        self.linear2 = nn.Linear(128,64)
        self.linear3 = nn.Linear(64,10)
    def forward(self,input):
        x = input.view(-1,28*28)
        x = nn.functional.relu(self.linear1(x))
        x = nn.functional.relu(self.linear2(x))
        y = self.linear3(x)
        return y

5 模型、损失函数、优化器设置

model = Model()
loss_fn  = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters(),lr=0.001)

6 精度、测试集精度和损失

def accuracy(y_pred,y_true):
    y_pred = torch.argmax(y_pred,dim=1)
    acc = (y_pred==y_true).float().mean()
    return acc
#测试集
def evaluate_testset(data_loader,model):
    acc_sum,loss_sum,total_example = 0.0,0.0,0
    for x,y in data_loader:
        y_hat = model(x)
        acc_sum += (y_hat.argmax(dim=1)==y).sum().item()
        loss = loss_fn(y_hat,y) 
        loss_sum += loss.item()
        total_example+=y.shape[0]
    return acc_sum/total_example,loss_sum

7 定义训练函数

#定义模型训练函数
def train(model,train_loader,test_loader,loss,num_epochs,batch_size,params=None,lr=None,optimizer=None):
    train_ls = []
    test_ls = []
    for epoch in range(num_epochs): # 训练模型一共需要num_epochs个迭代周期
        train_loss_sum, train_acc_num,total_examples = 0.0,0.0,0
        for x, y in train_loader: # x和y分别是小批量样本的特征和标签
            y_pred = model(x)
            loss = loss_fn(y_pred, y)  #计算损失
            optimizer.zero_grad() # 梯度清零
            loss.backward()  # 反向传播
            optimizer.step() #梯度更新
            total_examples += y.shape[0]
            train_loss_sum += loss.item()
            train_acc_num += (y_pred.argmax(dim=1)==y).sum().item()
        train_ls.append(train_loss_sum)
        test_acc,test_loss = evaluate_testset(test_loader,model)
        test_ls.append(test_loss)
        print('epoch %d, train_loss %.6f,test_loss %f,train_acc %.6f,test_acc %.6f'%(epoch+1, train_ls[epoch],test_ls[epoch],train_acc_num/total_examples,test_acc))
    return 

8 超参数设置

num_epoch  = 20
batch_size = 64

9 模型训练

train(model,train_loader,test_loader,loss_fn,num_epoch,batch_size,params=model.parameters,lr=0.001,optimizer=optimizer)