多输入模型

函数式API可以用于构建具有多个输入的模型，通常情况下，这种模型会在某一时刻用一个可以组合多个张量的层将不同的输入分支合并，张量组合方式可能是相加，连接等。这通常利用Keras的合并运算来实现，比如keras.layers.add, keras.layers.concatenate等。

下面来看一个非常简单的多输入模型示例：一个问答模型

典型的问答模型有两个输入，一个自然语言描述的问题和一个文本片段（比如新闻文章），后者提供用于回答问题的信息。然后模型要生成一个回答，在最简单的情况下，这个回答只包含一个词，可以通过对某个预定义的词表做softmax得到。

# 具有两个输入的模型
from keras.models import Model
from keras import layers
from keras import Input

text_vocabulary_size = 10000
question_vocabulary_size = 10000
answer_vocabulary_size = 500

text_input = Input(shape=(None, ), dtype='int32', name='text')

embedded_text = layers.Embedding(
                text_vocabulary_size, 64) (text_input)   # 将输入嵌入到长度为64的向量
encoded_text = layers.LSTM(32)(embedded_text)

# 对问题进行相同的处理，使用不同的层实例
question_input = Input(shape=(None, ), dtype='int32', name='question')  

embedded_question = layers.Embedding(
    question_vocabulary_size, 32)(question_input)
encoded_question = layers.LSTM(16)(embedded_question)

# 将编码后的问题和文本连接起来
concatenated = layers.concatenate([encoded_text, encoded_question], axis=-1) 

# 在上面添加一个softmax分类器
answer = layers.Dense(answer_vocabulary_size, activation='softmax')(concatenated)

model = Model([text_input, question_input], answer)
model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['acc'])

Using TensorFlow backend.
/usr/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:523: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
/usr/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:524: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint8 = np.dtype([("quint8", np.uint8, 1)])
/usr/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint16 = np.dtype([("qint16", np.int16, 1)])
/usr/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:526: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint16 = np.dtype([("quint16", np.uint16, 1)])
/usr/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:527: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint32 = np.dtype([("qint32", np.int32, 1)])
/usr/lib/python3.7/site-packages/tensorflow/python/framework/dtypes.py:532: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])

多输入模型#

多输入模型