Pytorch搭建网络训练葡萄酒分类数据集(三分类)

代码如下：

import torch
import torch.nn as nn
import torch.optim as optim
import torch.nn.functional as F
import torch.utils.data as Data
import numpy as np
from sklearn.datasets import load_wine

device=torch.device('cuda' if torch.cuda.is_available else 'cpu')#借助gpu训练


data = load_wine()#获取数据集
X = data.data
Y = data.target

x_train=torch.FloatTensor(X)#将数据集转化为tensor格式
y_train=torch.LongTensor(Y)

class MyDataset(Data.Dataset):#Dataset与Dataloader
    def __init__(self,x_train,y_train):
        self.x_train=x_train
        self.y_train=y_train
        
    def __getitem__(self,idx):
        return x_train[idx],y_train[idx]
    
    def __len__(self):
        return len(x_train)

train_dataset=MyDataset(x_train,y_train)
train_dataloader=Data.DataLoader(train_dataset,batch_size=16,shuffle=True)

class MyModel(nn.Module):#搭建模型
    def __init__(self):
        super(MyModel, self).__init__()
        self.liner1=nn.Linear(13,64)
        self.activate1=nn.ReLU()
        self.liner2=nn.Linear(64,8)
        self.activate2=nn.ReLU()
        self.liner3=nn.Linear(8,3)
    def forward(self,x):
        output=self.liner1(x)
        output=self.activate1(output)
        output = self.liner2(output)
        output = self.activate2(output)
        output = self.liner3(output)
        return output

model=MyModel().to(device)
optimizer=optim.Adam(model.parameters(),lr=0.001)
loss_fn=nn.CrossEntropyLoss()

Epoch=10000#没必要训练这么多轮
idx=0
for i in range(Epoch):#训练模型
    
    for x,y in train_dataloader:
        x,y=x.to(device),y.to(device)
        pred=model(x)
        loss=loss_fn(pred,y)
        idx+=1
        if idx%1000==0:
            print(loss)
        #三件套
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
        
import random #测试模型
success=0
cnt=1000
with torch.no_grad():#不产生梯度，表明当前计算不需要反向传播
    for i in range(cnt):
        t=random.randint(0,len(x_train)-1)
        x_test=x_train[t].to(device)
        y_test=y_train[t]
        pred=model(x_test)
#         print(pred)
        result=np.argmax(pred.cpu().detach().numpy())
        if result==y_test.cpu().detach().numpy():
            success+=1
print(success/cnt*100,'%')
#这里测试模型有个问题，就是数据是从训练集中随机找到的，不是那么合理

测试结果:

测试结果居然是100%，大概率是因为过拟合了