pytorch 模型保存的完整例子+pytorch 模型保存只保存可训练参数吗？是（+解决方案）

        测试使用的是一个liner model，还有更多的问题。pytorch 模型保存只保存可训练参数吗？

save模型

# 导入包
import glob
import os

import torch
import matplotlib.pyplot as plt
import random #用于数据迭代器生成随机数据

# 生成数据集 x1类别0，x2类别1
n_data = torch.ones(50, 2)  # 数据的基本形态
x1 = torch.normal(2 * n_data, 1)  # shape=(50, 2)
y1 = torch.zeros(50)  # 类型0 shape=(50, 1)
x2 = torch.normal(-2 * n_data, 1)  # shape=(50, 2)
y2 = torch.ones(50)  # 类型1 shape=(50, 1)
# 注意 x, y 数据的数据形式一定要像下面一样(torch.cat是合并数据)
x = torch.cat((x1, x2), 0).type(torch.FloatTensor)
y = torch.cat((y1, y2), 0).type(torch.FloatTensor)

# 数据集可视化
plt.scatter(x.data.numpy()[:, 0], x.data.numpy()[:, 1], c=y.data.numpy(), s=100, lw=0, cmap='RdYlGn')
plt.show()

# 数据读取：
def data_iter(batch_size, x, y):
    num_examples = len(x)
    indices = list(range(num_examples))
    random.shuffle(indices)  # 样本的读取顺序是随机的
    for i in range(0, num_examples, batch_size):
        j = torch.LongTensor(indices[i: min(i + batch_size, num_examples)]) #最后一次可能不足一个batch
        yield  x.index_select(0, j), y.index_select(0, j)

#############################################################################################################
def saver(model_state_dict, optimizer_state_dict, model_path, epoch, max_to_save=30):
    total_models = glob.glob(model_path + '*')
    if len(total_models) >= max_to_save:
        total_models.sort()
        os.remove(total_models[0])

    state_dict = {}
    state_dict["model_state_dict"] = model_state_dict
    state_dict["optimizer_state_dict"] = optimizer_state_dict

    torch.save(state_dict, model_path + 'h' + str(epoch))
    print('models {} save successfully!'.format(model_path + 'hahaha' + str(epoch)))



################################################################################################################

import torch.nn as nn
import torch.optim as optim



class net(nn.Module):
    def __init__(self, **kwargs):
        super(net, self).__init__(**kwargs)
        self.net = nn.Sequential(nn.Linear(2, 1), nn.ReLU())

    def forward(self, x):
        return self.net(x)

def loss(y_hat, y):
    return (y_hat - y.view(y_hat.size())) ** 2 / 2



def accuracy(y_hat, y):  #@save
    """计算预测正确的数量。"""
    cmp = y_hat.type(y.dtype) > 0.5 # 大于0.5类别1
    result=cmp.type(y.dtype)
    acc = 1-float(((result-y).sum())/ len(y))
    return acc;

lr = 0.03
num_epochs = 3 # 迭代次数
batch_size = 10 # 批量大小
model = net()
params =  list(model.parameters())
optimizer = torch.optim.Adam(params, 1e-4)

for epoch in range(num_epochs):
    for X, y_train in data_iter(batch_size, x, y):
        optimizer.zero_grad()
        l = loss(model(X), y_train).sum()  # l是有关小批量X和y的损失
        l.backward(retain_graph=True)
        optimizer.step()
        print(l)
    saver(model.state_dict(), optimizer.state_dict(), "./", epoch + 1,  max_to_save=100)

load模型

# 导入包
import glob
import os

import torch
import matplotlib.pyplot as plt
import random #用于数据迭代器生成随机数据

# 生成数据集 x1类别0，x2类别1
n_data = torch.ones(50, 2)  # 数据的基本形态
x1 = torch.normal(2 * n_data, 1)  # shape=(50, 2)
y1 = torch.zeros(50)  # 类型0 shape=(50, 1)
x2 = torch.normal(-2 * n_data, 1)  # shape=(50, 2)
y2 = torch.ones(50)  # 类型1 shape=(50, 1)
# 注意 x, y 数据的数据形式一定要像下面一样(torch.cat是合并数据)
x = torch.cat((x1, x2), 0).type(torch.FloatTensor)
y = torch.cat((y1, y2), 0).type(torch.FloatTensor)

# 数据集可视化
plt.scatter(x.data.numpy()[:, 0], x.data.numpy()[:, 1], c=y.data.numpy(), s=100, lw=0, cmap='RdYlGn')
plt.show()

# 数据读取：
def data_iter(batch_size, x, y):
    num_examples = len(x)
    indices = list(range(num_examples))
    random.shuffle(indices)  # 样本的读取顺序是随机的
    for i in range(0, num_examples, batch_size):
        j = torch.LongTensor(indices[i: min(i + batch_size, num_examples)]) #最后一次可能不足一个batch
        yield  x.index_select(0, j), y.index_select(0, j)

#############################################################################################################
def saver(model_state_dict, optimizer_state_dict, model_path, epoch, max_to_save=30):
    total_models = glob.glob(model_path + '*')
    if len(total_models) >= max_to_save:
        total_models.sort()
        os.remove(total_models[0])

    state_dict = {}
    state_dict["model_state_dict"] = model_state_dict
    state_dict["optimizer_state_dict"] = optimizer_state_dict

    torch.save(state_dict, model_path + 'h' + str(epoch))
    print('models {} save successfully!'.format(model_path + 'hahaha' + str(epoch)))



################################################################################################################

import torch.nn as nn
import torch.optim as optim



class net(nn.Module):
    def __init__(self, **kwargs):
        super(net, self).__init__(**kwargs)
        self.net = nn.Sequential(nn.Linear(2, 1), nn.ReLU())

    def forward(self, x):
        return self.net(x)

def loss(y_hat, y):
    return (y_hat - y.view(y_hat.size())) ** 2 / 2



def accuracy(y_hat, y):  #@save
    """计算预测正确的数量。"""
    cmp = y_hat.type(y.dtype) > 0.5 # 大于0.5类别1
    result=cmp.type(y.dtype)
    acc = 1-float(((result-y).sum())/ len(y))
    return acc;

lr = 0.03
num_epochs = 3 # 迭代次数
batch_size = 10 # 批量大小
model = net()
params =  list(model.parameters())
optimizer = torch.optim.Adam(params, 1e-4)

# for epoch in range(num_epochs):
#     for X, y_train in data_iter(batch_size, x, y):
#         optimizer.zero_grad()
#         l = loss(model(X), y_train).sum()  # l是有关小批量X和y的损失
#         l.backward(retain_graph=True)
#         optimizer.step()
#         print(l)
#     saver(model.state_dict(), optimizer.state_dict(), "./", epoch + 1,  max_to_save=100)




def loader(model_path):
    state_dict = torch.load(model_path)
    model_state_dict = state_dict["model_state_dict"]
    optimizer_state_dict = state_dict["optimizer_state_dict"]
    return model_state_dict, optimizer_state_dict

model_state_dict, optimizer_state_dict = loader("h1")
model.load_state_dict(model_state_dict)
optimizer.load_state_dict(optimizer_state_dict)

print('pretrained models loaded!')

pytorch 模型保存只保存可训练参数吗？是

class net(nn.Module):
    def __init__(self, **kwargs):
        super(net, self).__init__(**kwargs)
        self.net = nn.Sequential(nn.Linear(2, 1), nn.ReLU())
        self.notrain= torch.rand((64, 64), dtype=torch.float)

    def forward(self, x):
        return self.net(x)

解决方案

• 直接更改.data

class net(nn.Module):
    def __init__(self, **kwargs):
        super(net, self).__init__(**kwargs)
        self.net = nn.Sequential(nn.Linear(2, 1), nn.ReLU())
        # self.notrain = torch.rand((64, 64), dtype=torch.float)
        self.notrain = torch.nn.Parameter(torch.ones(64, 64))

    def forward(self, x):
        return self.net(x)

for epoch in range(num_epochs):
    for X, y_train in data_iter(batch_size, x, y):
        optimizer.zero_grad()
        l = loss(model(X), y_train).sum()  # l是有关小批量X和y的损失
        l.backward(retain_graph=True)
        optimizer.step()
        print(l)
        model.notrain.data = model.notrain.data+2
    saver(model.state_dict(), optimizer.state_dict(), "./", epoch + 1,  max_to_save=100)

TypeError: cannot assign ‘torch.cuda.FloatTensor’ as parameter ‘***’ (torch.nn.Parameter or None expected)

• self.weight = self.weight.detach()会报以上的错误，可以考虑使用

1. 在网络传播中detach(这种方法一般效率低)
2. 推荐注册为buffer，或者直接self.weight = torch.nn.Parameter(Tensor data, requires_grad = False)
3. model.*** = torch.nn.Parameter(torch.load("./SAVEPE.pt"))

参考与更多

PyTorch DataLoader的bug ：随机mask或者对数据的随机挑选产生的bug