.. _sec_mlp_concise:
多层感知机的简洁实现
====================
本节将介绍通过高级API更简洁地实现多层感知机。
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
from mxnet import gluon, init, npx
from mxnet.gluon import nn
from d2l import mxnet as d2l
npx.set_np()
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
import torch
from torch import nn
from d2l import torch as d2l
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
import tensorflow as tf
from d2l import tensorflow as d2l
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
import warnings
from d2l import paddle as d2l
warnings.filterwarnings("ignore")
import paddle
from paddle import nn
.. raw:: html
.. raw:: html
模型
----
与softmax回归的简洁实现( :numref:`sec_softmax_concise`\ )相比,
唯一的区别是我们添加了2个全连接层(之前我们只添加了1个全连接层)。
第一层是隐藏层,它包含256个隐藏单元,并使用了ReLU激活函数。
第二层是输出层。
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
net = nn.Sequential()
net.add(nn.Dense(256, activation='relu'),
nn.Dense(10))
net.initialize(init.Normal(sigma=0.01))
.. raw:: latex
\diilbookstyleoutputcell
.. parsed-literal::
:class: output
[07:09:43] ../src/storage/storage.cc:196: Using Pooled (Naive) StorageManager for CPU
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
net = nn.Sequential(nn.Flatten(),
nn.Linear(784, 256),
nn.ReLU(),
nn.Linear(256, 10))
def init_weights(m):
if type(m) == nn.Linear:
nn.init.normal_(m.weight, std=0.01)
net.apply(init_weights);
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
net = tf.keras.models.Sequential([
tf.keras.layers.Flatten(),
tf.keras.layers.Dense(256, activation='relu'),
tf.keras.layers.Dense(10)])
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
net = nn.Sequential(nn.Flatten(),
nn.Linear(784, 256),
nn.ReLU(),
nn.Linear(256, 10))
for layer in net:
if type(layer) == nn.Linear:
weight_attr = paddle.framework.ParamAttr(initializer=paddle.nn.initializer.Normal(mean=0.0, std=0.01))
layer.weight_attr = weight_attr
.. raw:: html
.. raw:: html
训练过程的实现与我们实现softmax回归时完全相同,
这种模块化设计使我们能够将与模型架构有关的内容独立出来。
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
batch_size, lr, num_epochs = 256, 0.1, 10
loss = gluon.loss.SoftmaxCrossEntropyLoss()
trainer = gluon.Trainer(net.collect_params(), 'sgd', {'learning_rate': lr})
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, trainer)
.. figure:: output_mlp-concise_f87756_33_0.svg
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
batch_size, lr, num_epochs = 256, 0.1, 10
loss = nn.CrossEntropyLoss(reduction='none')
trainer = torch.optim.SGD(net.parameters(), lr=lr)
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, trainer)
.. figure:: output_mlp-concise_f87756_36_0.svg
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
batch_size, lr, num_epochs = 256, 0.1, 10
loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)
trainer = tf.keras.optimizers.SGD(learning_rate=lr)
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, trainer)
.. figure:: output_mlp-concise_f87756_39_0.svg
.. raw:: html
.. raw:: html
.. raw:: latex
\diilbookstyleinputcell
.. code:: python
batch_size, lr, num_epochs = 256, 0.1, 10
loss = nn.CrossEntropyLoss(reduction='none')
trainer = paddle.optimizer.SGD(parameters=net.parameters(), learning_rate=lr)
train_iter, test_iter = d2l.load_data_fashion_mnist(batch_size)
d2l.train_ch3(net, train_iter, test_iter, loss, num_epochs, trainer)
.. figure:: output_mlp-concise_f87756_42_0.svg
.. raw:: html
.. raw:: html
小结
----
- 我们可以使用高级API更简洁地实现多层感知机。
- 对于相同的分类问题,多层感知机的实现与softmax回归的实现相同,只是多层感知机的实现里增加了带有激活函数的隐藏层。
练习
----
1. 尝试添加不同数量的隐藏层(也可以修改学习率),怎么样设置效果最好?
2. 尝试不同的激活函数,哪个效果最好?
3. 尝试不同的方案来初始化权重,什么方法效果最好?
.. raw:: html
.. raw:: html
`Discussions `__
.. raw:: html
.. raw:: html
`Discussions `__
.. raw:: html
.. raw:: html
`Discussions `__
.. raw:: html
.. raw:: html
`Discussions `__
.. raw:: html
.. raw:: html