前言
一、keras提供了三种定义模型的方式
1. 序列式(Sequential) API
序贯(sequential)API允许你为大多数问题逐层堆叠创建模型。虽然说对很多的应用来说，这样的一个手法很简单也解决了很多深度学习网络结构的构建，但是它也有限制－它不允许你创建模型有共享层或有多个输入或输出的网络。
2. 函数式(Functional) API
Keras函数式(functional)API为构建网络模型提供了更为灵活的方式。
它允许你定义多个输入或输出模型以及共享图层的模型。除此之外，它允许你定义动态(ad-hoc)的非周期性(acyclic)网络图。
模型是通过创建层的实例(layer instances)并将它们直接相互连接成对来定义的，然后定义一个模型(model)来指定那些层是要作为这个模型的输入和输出。
3.子类(Subclassing) API
补充知识：keras pytorch 构建模型对比
使用CIFAR10数据集，用三种框架构建Residual_Network作为例子，比较框架间的异同。
数据集格式
pytorch的数据集格式
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
|
import torchimport torch.nn as nnimport torchvision# Download and construct CIFAR-10 dataset.train_dataset = torchvision.datasets.CIFAR10(root='../../data/', train=True, download=True)# Fetch one data pair (read data from disk).image, label = train_dataset[0]print (image.size()) # torch.Size([3, 32, 32])print (label) # 6print (train_dataset.data.shape) # (50000, 32, 32, 3)# type(train_dataset.targets)==listprint (len(train_dataset.targets)) # 50000# Data loader (this provides queues and threads in a very simple way).train_loader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=64, shuffle=True)"""# 演示DataLoader返回的数据结构# When iteration starts, queue and thread start to load data from files.data_iter = iter(train_loader)# Mini-batch images and labels.images, labels = data_iter.next()print(images.shape) # torch.Size([100, 3, 32, 32])print(labels.shape) # torch.Size([100]) 可见经过DataLoader后，labels由list变成了pytorch内置的tensor格式"""# 一般使用的话是下面这种# Actual usage of the data loader is as below.for images, labels in train_loader: # Training code should be written here. pass |
keras的数据格式
|
1
2
3
4
5
6
|
import kerasfrom keras.datasets import cifar10(train_x, train_y) , (test_x, test_y) = cifar10.load_data()print(train_x.shape) # ndarray 类型： (50000, 32, 32, 3)print(train_y.shape) # (50000, 1) |
输入网络的数据格式不同
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
|
"""1: pytorch 都是内置torch.xxTensor输入网络，而keras的则是原生ndarray类型2: 对于multi-class的其中一种loss，即cross-entropy loss 而言， pytorch的api为 CorssEntropyLoss, 但y_true不能用one-hoe编码！这与keras，tensorflow 都不同。tensorflow相应的api为softmax_cross_entropy 他们的api都仅限于multi-class classification3*: 其实上面提到的api都属于categorical cross-entropy loss, 又叫 softmax loss，是函数内部先进行了 softmax 激活，再经过cross-entropy loss。 这个loss是cross-entropy loss的变种， cross-entropy loss又叫logistic loss 或 multinomial logistic loss。 实现这种loss的函数不包括激活函数，需要自定义。 pytorch对应的api为BCEloss(仅限于 binary classification), tensorflow 对应的api为 log_loss。 cross-entropy loss的第二个变种是 binary cross-entropy loss 又叫 sigmoid cross- entropy loss。 函数内部先进行了sigmoid激活，再经过cross-entropy loss。 pytorch对应的api为BCEWithLogitsLoss, tensorflow对应的api为sigmoid_cross_entropy"""# pytorchcriterion = nn.CrossEntropyLoss()...for epoch in range(num_epochs): for i, (images, labels) in enumerate(train_loader): images = images.to(device) labels = labels.to(device) # Forward pass outputs = model(images) # 对于multi-class cross-entropy loss # 输入y_true不需要one-hot编码 loss = criterion(outputs, labels)...# keras# 对于multi-class cross-entropy loss# 输入y_true需要one-hot编码train_y = keras.utils.to_categorical(train_y,10)...model.fit_generator(datagen.flow(train_x, train_y, batch_size=128), validation_data=[test_x,test_y], epochs=epochs,steps_per_epoch=steps_per_epoch, verbose=1)... |
整体流程
keras 流程
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
model = myModel()model.compile(optimizer=Adam(0.001),loss="categorical_crossentropy",metrics=["accuracy"])model.fit_generator(datagen.flow(train_x, train_y, batch_size=128), validation_data=[test_x,test_y], epochs=epochs,steps_per_epoch=steps_per_epoch, verbose=1, workers=4)#Evaluate the accuracy of the test datasetaccuracy = model.evaluate(x=test_x,y=test_y,batch_size=128)# 保存整个网络model.save("cifar10model.h5")"""# https://blog.csdn.net/jiandanjinxin/article/details/77152530# 使用# keras.models.load_model("cifar10model.h5")# 只保存architecture# json_string = model.to_json() # open('my_model_architecture.json','w').write(json_string) # 使用# from keras.models import model_from_json#model = model_from_json(open('my_model_architecture.json').read()) # 只保存weights# model.save_weights('my_model_weights.h5') #需要在代码中初始化一个完全相同的模型 # model.load_weights('my_model_weights.h5')#需要加载权重到不同的网络结构（有些层一样）中，例如fine-tune或transfer-learning，可以通过层名字来加载模型 # model.load_weights('my_model_weights.h5', by_name=True)""" |
pytorch 流程
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
|
model = myModel()# Loss and optimizercriterion = nn.CrossEntropyLoss()for epoch in range(num_epochs): for i, (images, labels) in enumerate(train_loader): images = images.to(device) labels = labels.to(device) # Forward pass outputs = model(images) loss = criterion(outputs, labels) # Backward and optimize # 将上次迭代计算的梯度值清0 optimizer.zero_grad() # 反向传播，计算梯度值 loss.backward() # 更新权值参数 optimizer.step() # model.eval()，让model变成测试模式，对dropout和batch normalization的操作在训练和测试的时候是不一样的# eval（）时，pytorch会自动把BN和DropOut固定住，不会取平均，而是用训练好的值。# 不然的话，一旦test的batch_size过小，很容易就会被BN层导致生成图片颜色失真极大。model.eval()with torch.no_grad(): correct = 0 total = 0 for images, labels in test_loader: images = images.to(device) labels = labels.to(device) outputs = model(images) _, predicted = torch.max(outputs.data, 1) total += labels.size(0) correct += (predicted == labels).sum().item() print('Accuracy of the model on the test images: {} %'.format(100 * correct / total))# Save the model checkpoint# 这是只保存了weightstorch.save(model.state_dict(), 'resnet.ckpt')"""# 使用# myModel.load_state_dict(torch.load('params.ckpt'))# 若想保存整个网络（architecture + weights)# torch.save(resnet, 'model.ckpt')# 使用#model = torch.load('model.ckpt')""" |
对比流程
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
|
#https://blog.csdn.net/dss_dssssd/article/details/83892824"""1: 准备数据(注意数据格式不同）2: 定义网络结构model3: 定义损失函数4: 定义优化算法 optimizer5: 训练-keras 5.1:编译模型（传入loss function和optimizer等） 5.2:训练模型（fit or fit_generator，传入数据)5: 训练-pytorch迭代训练： 5.1:准备好tensor形式的输入数据和标签(可选) 5.2:前向传播计算网络输出output和计算损失函数loss 5.3:反向传播更新参数 以下三句话一句也不能少： 5.3.1:将上次迭代计算的梯度值清0 optimizer.zero_grad() 5.3.2:反向传播，计算梯度值 loss.backward() 5.3.3:更新权值参数 optimizer.step()6: 在测试集上测试-keras model.evaluate6: 在测试集上测试-pytorch 遍历测试集，自定义metric7: 保存网络（可选） 具体实现参考上面代码""" |
构建网络
对比网络
1、对于keras，不需要input_channels，函数内部会自动获得，而pytorch则需要显示声明input_channels
2、对于pytorch Conv2d需要指定padding，而keras的则是same和valid两种选项（valid即padding=0）
3、keras的Flatten操作可以视作pytorch中的view
4、keras的dimension一般顺序是（H, W, C) (tensorflow 为backend的话），而pytorch的顺序则是（ C, H, W)
5、具体的变换可以参照下方，但由于没有学过pytorch，keras也刚入门，不能保证正确，日后学的更深入了之后再来看看。
pytorch 构建Residual-network
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
|
import torchimport torch.nn as nnimport torchvisionimport torchvision.transforms as transforms# Device configurationdevice = torch.device('cuda' if torch.cuda.is_available() else 'cpu')# Hyper-parametersnum_epochs = 80learning_rate = 0.001# Image preprocessing modulestransform = transforms.Compose([ transforms.Pad(4), transforms.RandomHorizontalFlip(), transforms.RandomCrop(32), transforms.ToTensor()])# CIFAR-10 dataset# train_dataset.data.shape#Out[31]: (50000, 32, 32, 3)# train_dataset.targets list# len(list)=5000train_dataset = torchvision.datasets.CIFAR10(root='./data/', train=True, transform=transform, download=True)test_dataset = torchvision.datasets.CIFAR10(root='../../data/', train=False, transform=transforms.ToTensor())# Data loadertrain_loader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=100, shuffle=True)test_loader = torch.utils.data.DataLoader(dataset=test_dataset, batch_size=100, shuffle=False)# 3x3 convolutiondef conv3x3(in_channels, out_channels, stride=1): return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=stride, padding=1, bias=False)# Residual blockclass ResidualBlock(nn.Module): def __init__(self, in_channels, out_channels, stride=1, downsample=None): super(ResidualBlock, self).__init__() self.conv1 = conv3x3(in_channels, out_channels, stride) self.bn1 = nn.BatchNorm2d(out_channels) self.relu = nn.ReLU(inplace=True) self.conv2 = conv3x3(out_channels, out_channels) self.bn2 = nn.BatchNorm2d(out_channels) self.downsample = downsample def forward(self, x): residual = x out = self.conv1(x) out = self.bn1(out) out = self.relu(out) out = self.conv2(out) out = self.bn2(out) if self.downsample: residual = self.downsample(x) out += residual out = self.relu(out) return out# ResNetclass ResNet(nn.Module): def __init__(self, block, layers, num_classes=10): super(ResNet, self).__init__() self.in_channels = 16 self.conv = conv3x3(3, 16) self.bn = nn.BatchNorm2d(16) self.relu = nn.ReLU(inplace=True) self.layer1 = self.make_layer(block, 16, layers[0]) self.layer2 = self.make_layer(block, 32, layers[1], 2) self.layer3 = self.make_layer(block, 64, layers[2], 2) self.avg_pool = nn.AvgPool2d(8) self.fc = nn.Linear(64, num_classes) def make_layer(self, block, out_channels, blocks, stride=1): downsample = None if (stride != 1) or (self.in_channels != out_channels): downsample = nn.Sequential( conv3x3(self.in_channels, out_channels, stride=stride), nn.BatchNorm2d(out_channels)) layers = [] layers.append(block(self.in_channels, out_channels, stride, downsample)) self.in_channels = out_channels for i in range(1, blocks): layers.append(block(out_channels, out_channels)) # [*[1,2,3]] # Out[96]: [1, 2, 3] return nn.Sequential(*layers) def forward(self, x): out = self.conv(x) # out.shape:torch.Size([100, 16, 32, 32]) out = self.bn(out) out = self.relu(out) out = self.layer1(out) out = self.layer2(out) out = self.layer3(out) out = self.avg_pool(out) out = out.view(out.size(0), -1) out = self.fc(out) return out model = ResNet(ResidualBlock, [2, 2, 2]).to(device)# pip install torchsummary or# git clone https://github.com/sksq96/pytorch-summaryfrom torchsummary import summary# input_size=(C,H,W)summary(model, input_size=(3, 32, 32))images,labels = iter(train_loader).next()outputs = model(images)# Loss and optimizercriterion = nn.CrossEntropyLoss()optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)# For updating learning ratedef update_lr(optimizer, lr): for param_group in optimizer.param_groups: param_group['lr'] = lr# Train the modeltotal_step = len(train_loader)curr_lr = learning_ratefor epoch in range(num_epochs): for i, (images, labels) in enumerate(train_loader): images = images.to(device) labels = labels.to(device) # Forward pass outputs = model(images) loss = criterion(outputs, labels) # Backward and optimize optimizer.zero_grad() loss.backward() optimizer.step() if (i+1) % 100 == 0: print ("Epoch [{}/{}], Step [{}/{}] Loss: {:.4f}" .format(epoch+1, num_epochs, i+1, total_step, loss.item())) # Decay learning rate if (epoch+1) % 20 == 0: curr_lr /= 3 update_lr(optimizer, curr_lr)# Test the modelmodel.eval()with torch.no_grad(): correct = 0 total = 0 for images, labels in test_loader: images = images.to(device) labels = labels.to(device) outputs = model(images) _, predicted = torch.max(outputs.data, 1) total += labels.size(0) correct += (predicted == labels).sum().item() print('Accuracy of the model on the test images: {} %'.format(100 * correct / total))# Save the model checkpointtorch.save(model.state_dict(), 'resnet.ckpt') |
keras 对应的网络构建部分
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
|
"""#pytorchdef conv3x3(in_channels, out_channels, stride=1): return nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=stride, padding=1, bias=False)"""def conv3x3(x,out_channels, stride=1): #out = spatial_2d_padding(x,padding=((1, 1), (1, 1)), data_format="channels_last") return Conv2D(filters=out_channels, kernel_size=[3,3], strides=(stride,stride),padding="same")(x)"""# pytorch# Residual blockclass ResidualBlock(nn.Module): def __init__(self, in_channels, out_channels, stride=1, downsample=None): super(ResidualBlock, self).__init__() self.conv1 = conv3x3(in_channels, out_channels, stride) self.bn1 = nn.BatchNorm2d(out_channels) self.relu = nn.ReLU(inplace=True) self.conv2 = conv3x3(out_channels, out_channels) self.bn2 = nn.BatchNorm2d(out_channels) self.downsample = downsample def forward(self, x): residual = x out = self.conv1(x) out = self.bn1(out) out = self.relu(out) out = self.conv2(out) out = self.bn2(out) if self.downsample: residual = self.downsample(x) out += residual out = self.relu(out) return out"""def ResidualBlock(x, out_channels, stride=1, downsample=False): residual = x out = conv3x3(x, out_channels,stride) out = BatchNormalization()(out) out = Activation("relu")(out) out = conv3x3(out, out_channels) out = BatchNormalization()(out) if downsample: residual = conv3x3(residual, out_channels, stride=stride) residual = BatchNormalization()(residual) out = keras.layers.add([residual,out]) out = Activation("relu")(out) return out"""#pytorchdef make_layer(self, block, out_channels, blocks, stride=1): downsample = None if (stride != 1) or (self.in_channels != out_channels): downsample = nn.Sequential( conv3x3(self.in_channels, out_channels, stride=stride), nn.BatchNorm2d(out_channels)) layers = [] layers.append(block(self.in_channels, out_channels, stride, downsample)) self.in_channels = out_channels for i in range(1, blocks): layers.append(block(out_channels, out_channels)) # [*[1,2,3]] # Out[96]: [1, 2, 3] return nn.Sequential(*layers)"""def make_layer(x, out_channels, blocks, stride=1): # tf backend: x.output_shape[-1]==out_channels #print("x.shape[-1] ",x.shape[-1]) downsample = False if (stride != 1) or (out_channels != x.shape[-1]): downsample = True out = ResidualBlock(x, out_channels, stride, downsample) for i in range(1, blocks): out = ResidualBlock(out, out_channels) return outdef KerasResidual(input_shape): images = Input(input_shape) out = conv3x3(images,16) # out.shape=(None, 32, 32, 16) out = BatchNormalization()(out) out = Activation("relu")(out) layer1_out = make_layer(out, 16, layers[0]) layer2_out = make_layer(layer1_out, 32, layers[1], 2) layer3_out = make_layer(layer2_out, 64, layers[2], 2) out = AveragePooling2D(pool_size=(8,8))(layer3_out) out = Flatten()(out) # pytorch 的nn.CrossEntropyLoss()会首先执行softmax计算 # 当换成keras时，没有tf类似的softmax_cross_entropy # 自带的categorical_crossentropy不会执行激活操作，因此得在Dense层加上activation out = Dense(units=10, activation="softmax")(out) model = Model(inputs=images,outputs=out) return modelinput_shape=(32, 32, 3)layers=[2, 2, 2]mymodel = KerasResidual(input_shape)mymodel.summary() |
pytorch model summary
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
|
---------------------------------------------------------------- Layer (type) Output Shape Param #================================================================ Conv2d-1 [-1, 16, 32, 32] 432 BatchNorm2d-2 [-1, 16, 32, 32] 32 ReLU-3 [-1, 16, 32, 32] 0 Conv2d-4 [-1, 16, 32, 32] 2,304 BatchNorm2d-5 [-1, 16, 32, 32] 32 ReLU-6 [-1, 16, 32, 32] 0 Conv2d-7 [-1, 16, 32, 32] 2,304 BatchNorm2d-8 [-1, 16, 32, 32] 32 ReLU-9 [-1, 16, 32, 32] 0 ResidualBlock-10 [-1, 16, 32, 32] 0 Conv2d-11 [-1, 16, 32, 32] 2,304 BatchNorm2d-12 [-1, 16, 32, 32] 32 ReLU-13 [-1, 16, 32, 32] 0 Conv2d-14 [-1, 16, 32, 32] 2,304 BatchNorm2d-15 [-1, 16, 32, 32] 32 ReLU-16 [-1, 16, 32, 32] 0 ResidualBlock-17 [-1, 16, 32, 32] 0 Conv2d-18 [-1, 32, 16, 16] 4,608 BatchNorm2d-19 [-1, 32, 16, 16] 64 ReLU-20 [-1, 32, 16, 16] 0 Conv2d-21 [-1, 32, 16, 16] 9,216 BatchNorm2d-22 [-1, 32, 16, 16] 64 Conv2d-23 [-1, 32, 16, 16] 4,608 BatchNorm2d-24 [-1, 32, 16, 16] 64 ReLU-25 [-1, 32, 16, 16] 0 ResidualBlock-26 [-1, 32, 16, 16] 0 Conv2d-27 [-1, 32, 16, 16] 9,216 BatchNorm2d-28 [-1, 32, 16, 16] 64 ReLU-29 [-1, 32, 16, 16] 0 Conv2d-30 [-1, 32, 16, 16] 9,216 BatchNorm2d-31 [-1, 32, 16, 16] 64 ReLU-32 [-1, 32, 16, 16] 0 ResidualBlock-33 [-1, 32, 16, 16] 0 Conv2d-34 [-1, 64, 8, 8] 18,432 BatchNorm2d-35 [-1, 64, 8, 8] 128 ReLU-36 [-1, 64, 8, 8] 0 Conv2d-37 [-1, 64, 8, 8] 36,864 BatchNorm2d-38 [-1, 64, 8, 8] 128 Conv2d-39 [-1, 64, 8, 8] 18,432 BatchNorm2d-40 [-1, 64, 8, 8] 128 ReLU-41 [-1, 64, 8, 8] 0 ResidualBlock-42 [-1, 64, 8, 8] 0 Conv2d-43 [-1, 64, 8, 8] 36,864 BatchNorm2d-44 [-1, 64, 8, 8] 128 ReLU-45 [-1, 64, 8, 8] 0 Conv2d-46 [-1, 64, 8, 8] 36,864 BatchNorm2d-47 [-1, 64, 8, 8] 128 ReLU-48 [-1, 64, 8, 8] 0 ResidualBlock-49 [-1, 64, 8, 8] 0 AvgPool2d-50 [-1, 64, 1, 1] 0 Linear-51 [-1, 10] 650================================================================Total params: 195,738Trainable params: 195,738Non-trainable params: 0----------------------------------------------------------------Input size (MB): 0.01Forward/backward pass size (MB): 3.63Params size (MB): 0.75Estimated Total Size (MB): 4.38---------------------------------------------------------------- |
keras model summary
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
|
__________________________________________________________________________________________________Layer (type) Output Shape Param # Connected to ==================================================================================================input_26 (InputLayer) (None, 32, 32, 3) 0 __________________________________________________________________________________________________conv2d_103 (Conv2D) (None, 32, 32, 16) 448 input_26[0][0] __________________________________________________________________________________________________batch_normalization_99 (BatchNo (None, 32, 32, 16) 64 conv2d_103[0][0] __________________________________________________________________________________________________activation_87 (Activation) (None, 32, 32, 16) 0 batch_normalization_99[0][0] __________________________________________________________________________________________________conv2d_104 (Conv2D) (None, 32, 32, 16) 2320 activation_87[0][0] __________________________________________________________________________________________________batch_normalization_100 (BatchN (None, 32, 32, 16) 64 conv2d_104[0][0] __________________________________________________________________________________________________activation_88 (Activation) (None, 32, 32, 16) 0 batch_normalization_100[0][0] __________________________________________________________________________________________________conv2d_105 (Conv2D) (None, 32, 32, 16) 2320 activation_88[0][0] __________________________________________________________________________________________________batch_normalization_101 (BatchN (None, 32, 32, 16) 64 conv2d_105[0][0] __________________________________________________________________________________________________add_34 (Add) (None, 32, 32, 16) 0 activation_87[0][0] batch_normalization_101[0][0] __________________________________________________________________________________________________activation_89 (Activation) (None, 32, 32, 16) 0 add_34[0][0] __________________________________________________________________________________________________conv2d_106 (Conv2D) (None, 32, 32, 16) 2320 activation_89[0][0] __________________________________________________________________________________________________batch_normalization_102 (BatchN (None, 32, 32, 16) 64 conv2d_106[0][0] __________________________________________________________________________________________________activation_90 (Activation) (None, 32, 32, 16) 0 batch_normalization_102[0][0] __________________________________________________________________________________________________conv2d_107 (Conv2D) (None, 32, 32, 16) 2320 activation_90[0][0] __________________________________________________________________________________________________batch_normalization_103 (BatchN (None, 32, 32, 16) 64 conv2d_107[0][0] __________________________________________________________________________________________________add_35 (Add) (None, 32, 32, 16) 0 activation_89[0][0] batch_normalization_103[0][0] __________________________________________________________________________________________________activation_91 (Activation) (None, 32, 32, 16) 0 add_35[0][0] __________________________________________________________________________________________________conv2d_108 (Conv2D) (None, 16, 16, 32) 4640 activation_91[0][0] __________________________________________________________________________________________________batch_normalization_104 (BatchN (None, 16, 16, 32) 128 conv2d_108[0][0] __________________________________________________________________________________________________activation_92 (Activation) (None, 16, 16, 32) 0 batch_normalization_104[0][0] __________________________________________________________________________________________________conv2d_110 (Conv2D) (None, 16, 16, 32) 4640 activation_91[0][0] __________________________________________________________________________________________________conv2d_109 (Conv2D) (None, 16, 16, 32) 9248 activation_92[0][0] __________________________________________________________________________________________________batch_normalization_106 (BatchN (None, 16, 16, 32) 128 conv2d_110[0][0] __________________________________________________________________________________________________batch_normalization_105 (BatchN (None, 16, 16, 32) 128 conv2d_109[0][0] __________________________________________________________________________________________________add_36 (Add) (None, 16, 16, 32) 0 batch_normalization_106[0][0] batch_normalization_105[0][0] __________________________________________________________________________________________________activation_93 (Activation) (None, 16, 16, 32) 0 add_36[0][0] __________________________________________________________________________________________________conv2d_111 (Conv2D) (None, 16, 16, 32) 9248 activation_93[0][0] __________________________________________________________________________________________________batch_normalization_107 (BatchN (None, 16, 16, 32) 128 conv2d_111[0][0] __________________________________________________________________________________________________activation_94 (Activation) (None, 16, 16, 32) 0 batch_normalization_107[0][0] __________________________________________________________________________________________________conv2d_112 (Conv2D) (None, 16, 16, 32) 9248 activation_94[0][0] __________________________________________________________________________________________________batch_normalization_108 (BatchN (None, 16, 16, 32) 128 conv2d_112[0][0] __________________________________________________________________________________________________add_37 (Add) (None, 16, 16, 32) 0 activation_93[0][0] batch_normalization_108[0][0] __________________________________________________________________________________________________activation_95 (Activation) (None, 16, 16, 32) 0 add_37[0][0] __________________________________________________________________________________________________conv2d_113 (Conv2D) (None, 8, 8, 64) 18496 activation_95[0][0] __________________________________________________________________________________________________batch_normalization_109 (BatchN (None, 8, 8, 64) 256 conv2d_113[0][0] __________________________________________________________________________________________________activation_96 (Activation) (None, 8, 8, 64) 0 batch_normalization_109[0][0] __________________________________________________________________________________________________conv2d_115 (Conv2D) (None, 8, 8, 64) 18496 activation_95[0][0] __________________________________________________________________________________________________conv2d_114 (Conv2D) (None, 8, 8, 64) 36928 activation_96[0][0] __________________________________________________________________________________________________batch_normalization_111 (BatchN (None, 8, 8, 64) 256 conv2d_115[0][0] __________________________________________________________________________________________________batch_normalization_110 (BatchN (None, 8, 8, 64) 256 conv2d_114[0][0] __________________________________________________________________________________________________add_38 (Add) (None, 8, 8, 64) 0 batch_normalization_111[0][0] batch_normalization_110[0][0] __________________________________________________________________________________________________activation_97 (Activation) (None, 8, 8, 64) 0 add_38[0][0] __________________________________________________________________________________________________conv2d_116 (Conv2D) (None, 8, 8, 64) 36928 activation_97[0][0] __________________________________________________________________________________________________batch_normalization_112 (BatchN (None, 8, 8, 64) 256 conv2d_116[0][0] __________________________________________________________________________________________________activation_98 (Activation) (None, 8, 8, 64) 0 batch_normalization_112[0][0] __________________________________________________________________________________________________conv2d_117 (Conv2D) (None, 8, 8, 64) 36928 activation_98[0][0] __________________________________________________________________________________________________batch_normalization_113 (BatchN (None, 8, 8, 64) 256 conv2d_117[0][0] __________________________________________________________________________________________________add_39 (Add) (None, 8, 8, 64) 0 activation_97[0][0] batch_normalization_113[0][0] __________________________________________________________________________________________________activation_99 (Activation) (None, 8, 8, 64) 0 add_39[0][0] __________________________________________________________________________________________________average_pooling2d_2 (AveragePoo (None, 1, 1, 64) 0 activation_99[0][0] __________________________________________________________________________________________________flatten_2 (Flatten) (None, 64) 0 average_pooling2d_2[0][0] __________________________________________________________________________________________________dense_2 (Dense) (None, 10) 650 flatten_2[0][0] ==================================================================================================Total params: 197,418Trainable params: 196,298Non-trainable params: 1,120__________________________________________________________________________________________________ |
以上这篇keras的三种模型实现与区别说明就是小编分享给大家的全部内容了，希望能给大家一个参考，也希望大家多多支持服务器之家。
原文链接：https://blog.csdn.net/yeziyezi1986/article/details/106780379
