深度学习框架Pytorch——学习笔记（三)线性回归

深度学习框架Pytorch——学习笔记（二)线性回归

使用pytorch 定义网络实现 线性回归

import torch
from torch.autograd import Variable
from matplotlib import pyplot as plt

# train data
# 训练数据 坐标位置(x,y)
x_data = Variable(torch.Tensor([[1.0], [2.0], [3.0]]))
y_data = Variable(torch.Tensor([[2.0], [4.0], [6.0]]))

# 定义网络 只用到了Linear线性映射
class Model(torch.nn.Module):
    def __init__(self):
        super(Model, self).__init__()
        self.linear = torch.nn.Linear(1, 1) 

    def forward(self, x):
        y_pred = self.linear(x)
        return y_pred
model = Model()

# 定义损失函数和优化方法
criterion = torch.nn.MSELoss(size_average=False) 
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)

# Training
for epoch in range(1000):
    # Forward pass
    y_pred = model(x_data)

    # Compute loss
    loss = criterion(y_pred, y_data)
    print(epoch, loss.data)

    # Zero gradients
    optimizer.zero_grad()
    # perform backward pass
    loss.backward()
    # update weights
    optimizer.step()

# 通过可视化验证训练结果
x = torch.arange(0, 20.0).view(-1, 1)
model.eval()
y = model(x)

plt.plot(x.data.numpy(), y.data.numpy())
plt.scatter(x_data.data.numpy(), y_data.data.numpy())

plt.xlim(0, 7)
plt.ylim(0, 7)
plt.show()
plt.pause(0.5)


# 输入一个值测试
var = Variable(torch.Tensor([4.0]))
print("predict (after training)", 4, model(var).data[0].numpy())

总结

1. 网络自定义 需要继承 torch.nn.Module 实现
2. torch.nn.Linear()  进行线性变换，**y =a*x**   通过多次迭代逐渐优化值，得到最佳a
3. net()的输入个数与输出个数相同，即维度相同