问题：
在keras中的预训练密集层之间添加dropout层
洪光霁
在keras中。应用程序中，有一个VGG16模型在imagenet上预先培训过。
from keras.applications import VGG16
model = VGG16(weights='imagenet')
该模型具有以下结构。
Layer (type) Output Shape Param # Connected to
====================================================================================================
input_1 (InputLayer) (None, 3, 224, 224) 0
____________________________________________________________________________________________________
block1_conv1 (Convolution2D) (None, 64, 224, 224) 1792 input_1[0][0]
____________________________________________________________________________________________________
block1_conv2 (Convolution2D) (None, 64, 224, 224) 36928 block1_conv1[0][0]
____________________________________________________________________________________________________
block1_pool (MaxPooling2D) (None, 64, 112, 112) 0 block1_conv2[0][0]
____________________________________________________________________________________________________
block2_conv1 (Convolution2D) (None, 128, 112, 112) 73856 block1_pool[0][0]
____________________________________________________________________________________________________
block2_conv2 (Convolution2D) (None, 128, 112, 112) 147584 block2_conv1[0][0]
____________________________________________________________________________________________________
block2_pool (MaxPooling2D) (None, 128, 56, 56) 0 block2_conv2[0][0]
____________________________________________________________________________________________________
block3_conv1 (Convolution2D) (None, 256, 56, 56) 295168 block2_pool[0][0]
____________________________________________________________________________________________________
block3_conv2 (Convolution2D) (None, 256, 56, 56) 590080 block3_conv1[0][0]
____________________________________________________________________________________________________
block3_conv3 (Convolution2D) (None, 256, 56, 56) 590080 block3_conv2[0][0]
____________________________________________________________________________________________________
block3_pool (MaxPooling2D) (None, 256, 28, 28) 0 block3_conv3[0][0]
____________________________________________________________________________________________________
block4_conv1 (Convolution2D) (None, 512, 28, 28) 1180160 block3_pool[0][0]
____________________________________________________________________________________________________
block4_conv2 (Convolution2D) (None, 512, 28, 28) 2359808 block4_conv1[0][0]
____________________________________________________________________________________________________
block4_conv3 (Convolution2D) (None, 512, 28, 28) 2359808 block4_conv2[0][0]
____________________________________________________________________________________________________
block4_pool (MaxPooling2D) (None, 512, 14, 14) 0 block4_conv3[0][0]
____________________________________________________________________________________________________
block5_conv1 (Convolution2D) (None, 512, 14, 14) 2359808 block4_pool[0][0]
____________________________________________________________________________________________________
block5_conv2 (Convolution2D) (None, 512, 14, 14) 2359808 block5_conv1[0][0]
____________________________________________________________________________________________________
block5_conv3 (Convolution2D) (None, 512, 14, 14) 2359808 block5_conv2[0][0]
____________________________________________________________________________________________________
block5_pool (MaxPooling2D) (None, 512, 7, 7) 0 block5_conv3[0][0]
____________________________________________________________________________________________________
flatten (Flatten) (None, 25088) 0 block5_pool[0][0]
____________________________________________________________________________________________________
fc1 (Dense) (None, 4096) 102764544 flatten[0][0]
____________________________________________________________________________________________________
fc2 (Dense) (None, 4096) 16781312 fc1[0][0]
____________________________________________________________________________________________________
predictions (Dense) (None, 1000) 4097000 fc2[0][0]
====================================================================================================
Total params: 138,357,544
Trainable params: 138,357,544
Non-trainable params: 0
____________________________________________________________________________________________________
我想用密集层（fc1、fc2和预测）之间的缺失层微调此模型，同时保持模型的所有预训练权重不变。我知道可以使用模型单独访问每一层。图层，但我还没有找到如何在现有图层之间添加新图层的方法。
这样做的最佳实践是什么？
共有2个答案
饶志
这是一个留在Keras“Sequential API”中的解决方案。
您可以在图层之间循环，并将其顺序添加到更新的顺序模型中。使用if子句在您选择的层之后添加退出项。
from tensorflow.keras.applications import VGG16
from tensorflow.keras.layers import Dropout
from tensorflow.keras.models import Sequential
model = VGG16(weights='imagenet')
# check structure and layer names before looping
model.summary()
# loop through layers, add Dropout after layers 'fc1' and 'fc2'
updated_model = Sequential()
for layer in model.layers:
updated_model.add(layer)
if layer.name in ['fc1', 'fc2']:
updated_model.add(Dropout(.2))
model = updated_model
# check structure
model.summary()
谭山
我自己通过使用Keras函数API找到了答案
from keras.applications import VGG16
from keras.layers import Dropout
from keras.models import Model
model = VGG16(weights='imagenet')
# Store the fully connected layers
fc1 = model.layers[-3]
fc2 = model.layers[-2]
predictions = model.layers[-1]
# Create the dropout layers
dropout1 = Dropout(0.85)
dropout2 = Dropout(0.85)
# Reconnect the layers
x = dropout1(fc1.output)
x = fc2(x)
x = dropout2(x)
predictors = predictions(x)
# Create a new model
model2 = Model(input=model.input, output=predictors)
Model2具有我想要的dropout层
____________________________________________________________________________________________________
Layer (type) Output Shape Param # Connected to
====================================================================================================
input_1 (InputLayer) (None, 3, 224, 224) 0
____________________________________________________________________________________________________
block1_conv1 (Convolution2D) (None, 64, 224, 224) 1792 input_1[0][0]
____________________________________________________________________________________________________
block1_conv2 (Convolution2D) (None, 64, 224, 224) 36928 block1_conv1[0][0]
____________________________________________________________________________________________________
block1_pool (MaxPooling2D) (None, 64, 112, 112) 0 block1_conv2[0][0]
____________________________________________________________________________________________________
block2_conv1 (Convolution2D) (None, 128, 112, 112) 73856 block1_pool[0][0]
____________________________________________________________________________________________________
block2_conv2 (Convolution2D) (None, 128, 112, 112) 147584 block2_conv1[0][0]
____________________________________________________________________________________________________
block2_pool (MaxPooling2D) (None, 128, 56, 56) 0 block2_conv2[0][0]
____________________________________________________________________________________________________
block3_conv1 (Convolution2D) (None, 256, 56, 56) 295168 block2_pool[0][0]
____________________________________________________________________________________________________
block3_conv2 (Convolution2D) (None, 256, 56, 56) 590080 block3_conv1[0][0]
____________________________________________________________________________________________________
block3_conv3 (Convolution2D) (None, 256, 56, 56) 590080 block3_conv2[0][0]
____________________________________________________________________________________________________
block3_pool (MaxPooling2D) (None, 256, 28, 28) 0 block3_conv3[0][0]
____________________________________________________________________________________________________
block4_conv1 (Convolution2D) (None, 512, 28, 28) 1180160 block3_pool[0][0]
____________________________________________________________________________________________________
block4_conv2 (Convolution2D) (None, 512, 28, 28) 2359808 block4_conv1[0][0]
____________________________________________________________________________________________________
block4_conv3 (Convolution2D) (None, 512, 28, 28) 2359808 block4_conv2[0][0]
____________________________________________________________________________________________________
block4_pool (MaxPooling2D) (None, 512, 14, 14) 0 block4_conv3[0][0]
____________________________________________________________________________________________________
block5_conv1 (Convolution2D) (None, 512, 14, 14) 2359808 block4_pool[0][0]
____________________________________________________________________________________________________
block5_conv2 (Convolution2D) (None, 512, 14, 14) 2359808 block5_conv1[0][0]
____________________________________________________________________________________________________
block5_conv3 (Convolution2D) (None, 512, 14, 14) 2359808 block5_conv2[0][0]
____________________________________________________________________________________________________
block5_pool (MaxPooling2D) (None, 512, 7, 7) 0 block5_conv3[0][0]
____________________________________________________________________________________________________
flatten (Flatten) (None, 25088) 0 block5_pool[0][0]
____________________________________________________________________________________________________
fc1 (Dense) (None, 4096) 102764544 flatten[0][0]
____________________________________________________________________________________________________
dropout_1 (Dropout) (None, 4096) 0 fc1[0][0]
____________________________________________________________________________________________________
fc2 (Dense) (None, 4096) 16781312 dropout_1[0][0]
____________________________________________________________________________________________________
dropout_2 (Dropout) (None, 4096) 0 fc2[1][0]
____________________________________________________________________________________________________
predictions (Dense) (None, 1000) 4097000 dropout_2[0][0]
====================================================================================================
Total params: 138,357,544
Trainable params: 138,357,544
Non-trainable params: 0
____________________________________________________________________________________________________
类似资料：
-
我在一维输入信号上实现了一个具有批量归一化的卷积神经网络。我的模型有一个相当好的精度约80%。以下是我的图层顺序：（Conv1D，Batch，ReLU，MaxPooling）重复6次，Conv1D，Batch，ReLU，Dense，Softmax。 我看过几篇文章说我不应该在卷积层上使用dropout，而是应该使用批处理规范化，所以我想通过用dropout层替换所有批处理规范化层来试验我的模型，看
-
数据增强采用的方法是从现有的示例中生成额外的训练数据，方法是通过增强，然后使用随机转换来生成看起来可信的图像。这有助于将模型公开给数据的更多方面，并更好地泛化。 所以我对此的理解是--例如，如果我没有太多的训练图像--我希望通过在现有的训练图像之外创建新的、增强的图像来生成额外的训练数据。 然后在上面链接的Keras文档中显示了如何将模块中的一些预处理层作为第一层添加到示例的模型中。所以从理论上讲
-
文章信息 通过本教程，你可以掌握技能：使用预先训练的词向量和卷积神经网络解决一个文本分类问题 本文代码已上传到Github 本文地址：http://blog.keras.io/using-pre-trained-word-embeddings-in-a-keras-model.html 本文作者：Francois Chollet 什么是词向量? ”词向量”（词嵌入）是将一类将词的语义映射到向量空间
-
问题内容： 我想知道是否有可能保存经过部分训练的Keras模型并在再次加载模型后继续进行训练。 这样做的原因是，将来我将拥有更多的训练数据，并且我不想再次对整个模型进行训练。 我正在使用的功能是： 编辑1：添加了完全正常的示例 对于10个纪元后的第一个数据集，最后一个纪元的损失将为0.0748，精度为0.9863。 保存，删除和重新加载模型后，第二个数据集上训练的模型的损失和准确性分别为0.171
-
问题内容： 因此，我一直遵循Google的官方tensorflow指南，并尝试使用Keras构建一个简单的神经网络。但是，在训练模型时，它不使用整个数据集（具有60000个条目），而是仅使用1875个条目进行训练。有可能解决吗？ 输出： 这是我一直在为此工作的原始Google colab笔记本：https ://colab.research.google.com/drive/1NdtzXHEpiN
-
我想知道是否有可能在卷积神经网络的密集层中添加一个变量（以及之前卷积层的连接，是否有额外的特征集可用于区分目的）？如果可能的话，有人能给我举个例子/文档来解释如何做到这一点吗？ 我希望使用Keras，但如果Keras限制太多，我很乐意使用TensorFlow。 编辑：在这种情况下，我认为这应该起作用的方式是我向神经网络提供一个包含图像和相关特征集的列表（以及在训练相关分类期间）。 EDIT2：我想