源码:github code
pytorch线性回归的例子
import torch
import torch.nn as nn
import numpy as np
import matplotlib.pyplot as plt
# Hyper-parameters
# 超参设置
input_size = 1
output_size = 1
num_epochs = 60
learning_rate = 0.001
# Toy dataset
# 数据
x_train = np.array([[3.3], [4.4], [5.5], [6.71], [6.93], [4.168],
[9.779], [6.182], [7.59], [2.167], [7.042],
[10.791], [5.313], [7.997], [3.1]], dtype=np.float32)
y_train = np.array([[1.7], [2.76], [2.09], [3.19], [1.694], [1.573],
[3.366], [2.596], [2.53], [1.221], [2.827],
[3.465], [1.65], [2.904], [1.3]], dtype=np.float32)
# Linear regression model
# 线性回归模型
model = nn.Linear(input_size, output_size)
print(model.weight)
print(model.bias)
# Loss and optimizer
# 损失函数和优化器
criterion = nn.MSELoss()
optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)
# Train the model
# 模型训练
for epoch in range(num_epochs):
# Convert numpy arrays to torch tensors
# numpy数据转换为tensor数据
inputs = torch.from_numpy(x_train)
targets = torch.from_numpy(y_train)
# Forward pass
outputs = model(inputs)
loss = criterion(outputs, targets)
# Backward and optimize
optimizer.zero_grad()
loss.backward()
optimizer.step()
if (epoch + 1) % 5 == 0:
print('Epoch [{}/{}], Loss: {:.4f}'.format(epoch + 1, num_epochs, loss.item()))
# Plot the graph
# 画图--线性回归
predicted = model(torch.from_numpy(x_train)).detach().numpy()
plt.plot(x_train, y_train, 'ro', label='Original data')
plt.plot(x_train, predicted, 'yo', label='Predict data')
plt.plot(x_train, predicted, label='Fitted line')
plt.legend()
plt.show()
# Save the model checkpoint
# 保存模型
torch.save(model.state_dict(), 'model.ckpt')
Parameter containing:
tensor([[-0.9154]], requires_grad=True)
Parameter containing:
tensor([0.8025], requires_grad=True)
Epoch [5/60], Loss: 28.1747
Epoch [10/60], Loss: 11.5182
Epoch [15/60], Loss: 4.7704
Epoch [20/60], Loss: 2.0368
Epoch [25/60], Loss: 0.9293
Epoch [30/60], Loss: 0.4806
Epoch [35/60], Loss: 0.2989
Epoch [40/60], Loss: 0.2252
Epoch [45/60], Loss: 0.1954
Epoch [50/60], Loss: 0.1833
Epoch [55/60], Loss: 0.1784
Epoch [60/60], Loss: 0.1764