PyTorch深度学习模型的保存和加载流程详解

import torch
import torch.nn as nn

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 2, 3)
        self.pool1 = nn.MaxPool2d(2, 2)

    def forward(self, x):
        x = self.conv1(x)
        x = self.pool1(x)
        return x

# 初始化网络
net = Net()
net.conv1.weight[0].detach().fill_(1)
net.conv1.weight[1].detach().fill_(2)
net.conv1.bias.data.detach().zero_()
# 获取state_dict
state_dict = net.state_dict()
# 字典的遍历默认是遍历key，所以param_tensor实际上是键值
for param_tensor in state_dict: 
    print(param_tensor,'：\n',state_dict[param_tensor])
# 保存模型参数
torch.save(state_dict,"net_params.pth")
# 通过加载state_dict获取模型参数
net.load_state_dict(state_dict)

import torch
import torch.nn as nn

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 2, 3)
        self.pool1 = nn.MaxPool2d(2, 2)

    def forward(self, x):
        x = self.conv1(x)
        x = self.pool1(x)
        return x

# 初始化网络
net = Net()
net.conv1.weight[0].detach().fill_(1)
net.conv1.weight[1].detach().fill_(2)
net.conv1.bias.data.detach().zero_()
# 保存整个网络
torch.save(net,"net.pth")
# 加载网络
net = torch.load("net.pth")