model.parameters()，model.state_dict()，model .load_state_dict()以及torch.load()

淡淡的烟草味﹌ 2023-10-04 16:59 71阅读 0赞

## 一.model.parameters()与model.state\_dict() ##

*model.parameters()*与*model.state\_dict()都*是Pytorch中用于查看网络参数的方法

一般来说，前者多见于优化器的初始化，例如：

![2bd8e1aabfc19a96d15a13042c969381.png][]

后者多见于模型的保存，如：

![dc9b75edaf55e4e8777399612a7e156c.png][]

当我们对网络调参或者查看网络的参数是否具有可复现性时，可能会查看网络的参数

pretrained_dict = torch.load(yolov4conv137weight)
    
    model_dict = _model.state_dict()  #查看模型的权重和biass系数
              
    pretrained_dict = {k1: v for (k, v), k1 in zip(pretrained_dict.items(), model_dict)}
    
    model_dict.update(pretrained_dict) #更新model网络模型的参数的权值和biass，这相当于是一个浅拷贝，对这个更新改变会更改模型的权重和biass

`model.state_dict()`其实返回的是一个`OrderDict`，存储了网络结构的名字和对应的参数。

例子：

#encoding:utf-8
     
    import torch
    import torch.nn as nn
    import torch.optim as optim
    import torchvision
    import numpy as mp
    import matplotlib.pyplot as plt
    import torch.nn.functional as F
     
    #define model
    class TheModelClass(nn.Module):
        def __init__(self):
            super(TheModelClass,self).__init__()
            self.conv1=nn.Conv2d(3,6,5)
            self.pool=nn.MaxPool2d(2,2)
            self.conv2=nn.Conv2d(6,16,5)
            self.fc1=nn.Linear(16*5*5,120)
            self.fc2=nn.Linear(120,84)
            self.fc3=nn.Linear(84,10)
     
        def forward(self,x):
            x=self.pool(F.relu(self.conv1(x)))
            x=self.pool(F.relu(self.conv2(x)))
            x=x.view(-1,16*5*5)
            x=F.relu(self.fc1(x))
            x=F.relu(self.fc2(x))
            x=self.fc3(x)
            return x
     
    def main():
        # Initialize model
        model = TheModelClass()
     
        #Initialize optimizer
        optimizer=optim.SGD(model.parameters(),lr=0.001,momentum=0.9)
     
        #print model's state_dict
        print('Model.state_dict:')
        for param_tensor in model.state_dict():
            #打印 key value字典
            print(param_tensor,'\t',model.state_dict()[param_tensor].size())
     
        #print optimizer's state_dict
        print('Optimizer,s state_dict:')
        for var_name in optimizer.state_dict():
            print(var_name,'\t',optimizer.state_dict()[var_name])
     
     
     
    if __name__=='__main__':
        main()

具体的**输出结果如下**：可以很清晰的观测到state\_dict中存放的key和value的值

Model.state_dict:
    conv1.weight 	 torch.Size([6, 3, 5, 5])
    conv1.bias 	 torch.Size([6])
    conv2.weight 	 torch.Size([16, 6, 5, 5])
    conv2.bias 	 torch.Size([16])
    fc1.weight 	 torch.Size([120, 400])
    fc1.bias 	 torch.Size([120])
    fc2.weight 	 torch.Size([84, 120])
    fc2.bias 	 torch.Size([84])
    fc3.weight 	 torch.Size([10, 84])
    fc3.bias 	 torch.Size([10])
    Optimizer,s state_dict:
    state 	 {}
    param_groups 	 [{'lr': 0.001, 'momentum': 0.9, 'dampening': 0, 'weight_decay': 0, 'nesterov': False, 'params': [367949288, 367949432, 376459056, 381121808, 381121952, 381122024, 381121880, 381122168, 381122096, 381122312]}]

## 二.torch.load()和load\_state\_dict() ##

load\_state\_dict(state\_dict, strict=True)

从 state\_dict 中复制参数和缓冲区到 Module 及其子类中

state\_dict：包含参数和缓冲区的 Module 状态字典

strict：默认 True，是否严格匹配 state\_dict 的键值和 Module.state\_dict()的键值

model = nn.Sequential(self.down1, self.down2, self.down3, self.down4, self.down5, self.neek)
    
    pretrained_dict = torch.load(yolov4conv137weight)  #加载已经训练好的模型参数
    
    model_dict = model.state_dict()  #查看权重和偏重
    
    # 1. filter out unnecessary keys
    pretrained_dict = {k1: v for (k, v), k1 in zip(pretrained_dict.items(), model_dict)}
    
    # 2. overwrite entries in the existing state dict
     model_dict.update(pretrained_dict)  #更新已有的模型的权重和偏重
      
    model.load_state_dict(model_dict)   #将更新后的参数重新加载至网络模型中

**官方推荐的方法，只保存和恢复模型中的参数**

# save
    torch.save(model.state_dict(), PATH)
     
    # load
    model = MyModel(*args, **kwargs)
    model.load_state_dict(torch.load(PATH))
    model.eval()

torch.load("path路径")表示加载已经训练好的模型

而model.load\_state\_dict（torch.load(PATH)）表示将训练好的模型参数重新加载至网络模型中

[2bd8e1aabfc19a96d15a13042c969381.png]: https://img-blog.csdnimg.cn/img_convert/2bd8e1aabfc19a96d15a13042c969381.png
[dc9b75edaf55e4e8777399612a7e156c.png]: https://img-blog.csdnimg.cn/img_convert/dc9b75edaf55e4e8777399612a7e156c.png