PyTorch深度学习实践（03.Gradient Descent）

源自课程：《PyTorch深度学习实践》完结合集

Chapter3 梯度下降算法

"""
梯度下降算法的基本实现
"""

x_data = [1.0, 2.0, 3.0]
y_data = [2.0, 4.0, 6.0]

# 对参数w进行初始化
w = 1.0


# 前向传播
def forward(x):
    return x*w


# 损失函数
def loss(x, y):
    y_pred = forward(x)
    return (y_pred - y)**2


# 梯度下降算法，计算梯度值
def gradient(x, y):
    # 手动求导计算导数
    return 2 * x * (x*w - y)


# 训练前的预测值
print('Predict (before training)', 4, forward(4))

for epoch in range(100):
    for x, y in zip(x_data, y_data):
        grad = gradient(x, y)
        # 设置参数更新速率为0.01
        w = w - 0.01 * grad
        print("\tgrad: ", x, y, grad)
        l = loss(x, y)

    print("progress:", epoch, "w=", w, "loss=", l)

# 训练后的预测值
print('Predict (after training)', 4, forward(4))