[人工智能]《PyTorch深度学习实践》 学习笔记2

1、反向传播

????????先进行前馈运算（forward），然后反向传播算出损失函数对权重的倒数（即梯度），进而可以进行更新权重w。

?

?

import torch

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

w = torch.Tensor([1.0])

w.requires_grad = True  # 计算梯度


def forward(x):
    return x * w


def loss(x, y):
    y_pred = forward(x)
    return (y_pred - y) ** 2


print('predict (before tarining)', 4, forward(4).item())

for i in range(100):
    for x, y in zip(x_data, y_data):
        ls = loss(x, y)
        ls.backward()
        w.data=w.data-0.01*w.grad.data
        w.grad.data.zero_()
    print("progress:",i,ls.item())
print('predict (after training)',4,forward(4).item())