PyTorch深度学习笔记（二十）（模型验证测试）

希露菲叶特格雷拉特2025-11-08 12:26

前言

到这一章节为止，依据小土堆课程的PyTorch深度学习笔记基础部分已经完结了，接下来将依据李沐动手学深度学习课程进行PyTorch深度学习笔记的进阶部分

预测图片

完整的模型验证(测试，demo)套路，利用已经训练好的模型，然后给它提供输入。

输入狗的图片，并打开

python 复制代码

image_path = "imgs/dog.png"
image = Image.open(image_path)

4通道的RGBA转为3通道的RGB图片

python 复制代码

image = image.convert("RGB")

转换图像格式并设定网络

python 复制代码

transform = torchvision.transforms.Compose([torchvision.transforms.Resize((32,32)),   
                                            torchvision.transforms.ToTensor()])

image = transform(image)
print(image.shape)

class Tudui(nn.Module):
    def __init__(self):
        super(Tudui, self).__init__()        
        self.model1 = 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.model1(x)
        return x

GPU上训练的东西映射到CPU上

python 复制代码

model = torch.load("model/tudui_29.pth",map_location=torch.device('cpu'))

转为四维，符合网络输入需求

python 复制代码

image = torch.reshape(image,(1,3,32,32))

将模型转为测试类型

python 复制代码

model.eval()

不进行梯度计算，减少内存计算

python 复制代码

with torch.no_grad():
    output = model(image)

概率最大类别的输出

python 复制代码

print(output.argmax(1))

完整代码

python 复制代码

import torchvision
from PIL import Image
from torch import nn
import torch

image_path = "imgs/dog.png"
image = Image.open(image_path)
image = image.convert("RGB") 
print(image)

transform = torchvision.transforms.Compose([torchvision.transforms.Resize((32,32)),   
                                            torchvision.transforms.ToTensor()])

image = transform(image)
print(image.shape)

class Tudui(nn.Module):
    def __init__(self):
        super(Tudui, self).__init__()        
        self.model1 = 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.model1(x)
        return x

model = torch.load("model/tudui_29.pth",map_location=torch.device('cpu'))
print(model)
image = torch.reshape(image,(1,3,32,32)) 
model.eval()
with torch.no_grad():
    output = model(image)
output = model(image)
print(output)
print(output.argmax(1))