学习笔记：动手学深度学习 14 多层感知机的从零开始实现

import torch
from torch import nn
from d2l import torch as d2l

#数据导入
batch_size = 256
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)
"""初始化模型参数"""
""" 输入输出管不了，256为超参数"""
num_inputs, num_outputs, num_hiddens = 784, 10, 256
"""权重初始化为随机的，偏差设为0"""
W1 = nn.Parameter(torch.randn(
    num_inputs, num_hiddens, requires_grad=True) * 0.01)
b1 = nn.Parameter(torch.zeros(num_hiddens, requires_grad=True))
W2 = nn.Parameter(torch.randn(
    num_hiddens, num_outputs, requires_grad=True) * 0.01)
b2 = nn.Parameter(torch.zeros(num_outputs, requires_grad=True))
params = [W1, b1, W2, b2]
"""激活函数，使用最大值函数自己实现ReLU激活函数"""
def relu(X):
    a = torch.zeros_like(X)
    return torch.max(X, a)
"""模型"""
def net(X):
    #-1就是bachsize
    X = X.reshape((-1, num_inputs))
    H = relu(X@W1 + b1)  # 这里“@”代表矩阵乘法
    return (H@W2 + b2)
"""损失函数"""
loss = nn.CrossEntropyLoss()
"""定义优化算法"""
def sgd(params, lr, batch_size):  #@save
    """params给定的list，小批量随机梯度下降。"""
    with torch.no_grad():