用model实现一个简单网络的例子

网络结构在文章https://blog.csdn.net/Acmer_future_victor/article/details/103490422中有。

# __author: Y
# date: 2019/12/11

import numpy as np
import torch
import torch.nn as nn

# N是有多少个训练数据 D_in是输入维度
N , D_in, H, D_out = 64, 1000, 100, 10
# 随机创建一些训练数据
x = torch.randn(N, D_in)
y = torch.randn(N, D_out)
learning_rate = 1e-4

#定义一个网络模型
model = nn.Sequential(
    nn.Linear(D_in, H, bias=False),  # w_1*x+b_1  多了一个b_1
    nn.ReLU(N, D_out, bias=False),
    nn.Linear(H, D_out, bias=False)
)

# 权重初始化，为了得到一个好的优化结果
nn.init.normal_(model[0].weight)
nn.init.normal_(model[2].weight)

# model = model.cuda()
#损失函数
loss_fn = nn.MSELoss(reduction='sum')
#优化器
optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)

for it in range(500):
    # Forward pass
    y_pred = model(x)  # model.forward()

    # compute loss
    loss = loss_fn(y_pred, y)  # computation graph
    print(it, loss.item())

    optimizer.zero_grad()
    # Backward pass
    # compute the gradient
    loss.backward()  # 这一句就做了上面所有的操作

    # update weights of w1 and w2
    # with torch.no_grad():
    #     for param in model.parameters():
    #         param -= learning_rate * param.grad
    # model.zero_grad()  # 防止param的grad累加，下一次backward之前清就可以

    optimizer.step()