动手学深度学习适合哪些人？MXNet 代码示例 + 在线调试全解析

import mxnet as mx  
from mxnet import autograd, gluon, init, nd  
from mxnet.gluon import data as gdata, nn  

mnist_train = gdata.vision.MNIST(train=True)  
mnist_test = gdata.vision.MNIST(train=False)  

def transform(data, label):  
    return data.astype('float32') / , label.astype('float32')  

batch_size =   
train_iter = gdata.DataLoader(mnist_train.transform(transform), batch_size, shuffle=True)  
test_iter = gdata.DataLoader(mnist_test.transform(transform), batch_size, shuffle=False)  

net = nn.Sequential()  
net.add(nn.Conv2D(channels=, kernel_size=, activation='relu'),  
        nn.MaxPool2D(pool_size=, strides=),  
        nn.Conv2D(channels=, kernel_size=, activation='relu'),  
        nn.MaxPool2D(pool_size=, strides=),  
        nn.Flatten(),  
        nn.Dense(, activation='relu'),  
        nn.Dense(, activation='relu'),  
        nn.Dense())  

net.initialize(init=init.Xavier())  

loss = gluon.loss.SoftmaxCrossEntropyLoss()  
trainer = gluon.Trainer(net.collect_params(), 'sgd', {'learning_rate': 0.1})  

num_epochs =   
for epoch in range(num_epochs):  
    for X, y in train_iter:  
        with autograd.record():  
            y_hat = net(X)  
            l = loss(y_hat, y)  
        l.backward()  
        trainer.step(batch_size)  
    test_acc = evaluate_accuracy(test_iter, net)  
    print(f'Epoch {epoch + }, Test Acc: {test_acc:.2f}')  

def evaluate_accuracy(iter, net):  
    acc = nd.array([])  
    for X, y in iter:  
        acc += (net(X).argmax(axis=) == y).sum()  
    return acc.asscalar() / len(iter.dataset)