softmax回归和多层感知机的简洁实现

softmax回归

import d2lzh as d2l
from mxnet import gluon, init
from mxnet.gluon import loss as gloss, nn

# 读取数据集
batch_size = 256
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)

# 定义和初始化模型
net = nn.Sequential()
net.add(nn.Dense(10))
net.initialize(init.Normal(sigma=0.01))

# softmax和交叉熵函数
loss = gloss.SoftmaxCrossEntropyLoss()

# 定义优化算法
trainer = gluon.Trainer(net.collect_params(), 'sgd', {'learning_rate':0.1})

# 训练模型
num_epochs = 5
# train_ch3为一个训练函数
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, batch_size, None, None, trainer)

多层感知机

%matplotlib inline
import d2lzh as d2l
from mxnet import nd, init
from mxnet.gluon import loss as gloss, nn, Trainer

batch_size = 256
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)

# 只是模型的定义不一样，其他地方的步骤都一致
net = nn.Sequential()
net.add(nn.Dense(256, activation='relu'),
       nn.Dense(10))
net.initialize(init.Normal(sigma=0.01))

loss = gloss.SoftmaxCrossEntropyLoss()
trainer = Trainer(net.collect_params(),'sgd',{'learning_rate':0.5})

# 开始训练
num_epochs = 5
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, batch_size, None, None, trainer)