Pytorch 深度学习实战（三）

Gradient Descent

# 梯度下降法

# 1. 准备数据集
x_data = [1.0, 2.0, 3.0]
y_data = [2.0, 4.0, 6.0]

# 初始化权重
w = 1.0

# 2. 定义网络
def forward(x):
    return x * w

# 3. 构建损失函数
def loss(xs, ys):
    cost = 0
    for x, y in zip(xs, ys):
        y_pred = forward(x)
        cost += (y_pred - y)  ** 2
    return cost / len(xs)

# 计算梯度
def gradient(xs, ys):
    grad = 0
    for x, y in zip(xs, ys):
        grad += 2 * x * (x * w - y)
    return grad / len(xs)

print('Predict (Before training)', 4 ,forward(4))

# 4. 训练循环
for epoch in range(100):
    # 前向传播
    cost_val = loss(x_data, y_data)
    # 反向传播
    grad_val = gradient(x_data, y_data)
    # 更新参数
    w -= 0.01 * grad_val
    
    print('Epoch:', epoch, 'w = ', w, 'loss = ', cost_val)

print('Predict (After training)', 4, forward(4))