1.简介

Seq2Seq技术，全称Sequence to Sequence，该技术突破了传统的固定大小输入问题框架，开通了将经典深度神经网络模型（DNNs）运用于在翻译，文本自动摘要和机器人自动问答以及一些回归预测任务上,并被证实在英语－法语翻译、英语－德语翻译以及人机短问快答的应用中有着不俗的表现。

2.核心思想

Seq2Seq解决问题的主要思路是通过深度神经网络模型（常用的是LSTM，长短记忆网络，一种循环神经网络）http://dataxujing.coding.me/深度学习之RNN/。将一个作为输入的序列映射为一个作为输出的序列，这一过程由编码（Encoder）输入与解码（Decoder）输出两个环节组成, 前者负责把序列编码成一个固定长度的向量，这个向量作为输入传给后者，输出可变长度的向量。

3.代码实现

(1)训练代码

import numpy as np
import tensorflow as tf
from tensorflow import keras
from keras import layers
batch_size=64
epochs=30
latent_dim=128
num_samples=10000
data_path='../dataset/fra-eng/fra.txt'


#prepare data
input_texts=[]
target_texts=[]
input_characters=set()
target_characters=set()

with open(data_path, 'r', encoding='utf-8') as f:
    lines=f.read().split('\n')

for line in lines[: min(num_samples, len(lines) - 1)]:
    input_text, target_text, _ = line.split("\t")
    # We use "tab" as the "start sequence" character
    # for the targets, and "\n" as "end sequence" character.
    target_text = "\t" + target_text + "\n"
    input_texts.append(input_text)
    target_texts.append(target_text)
    for char in input_text:
        if char not in input_characters:
            input_characters.add(char)
    for char in target_text:
        if char not in target_characters:
            target_characters.add(char)

input_characters=sorted(list(input_characters))
target_characters=sorted(list(target_characters))
num_encoder_tokens=len(input_characters)
num_decoder_tokens=len(target_characters)
max_encoder_seq_length=max([len(txt) for txt in input_texts])
max_decoder_seq_length=max([len(txt) for txt in target_texts])

input_token_index = dict([(char, i) for i, char in enumerate(input_characters)])
target_token_index = dict([(char, i) for i, char in enumerate(target_characters)])


# print(input_token_index)
# print(target_token_index)

encoder_input_data = np.zeros(
    (len(input_texts), max_encoder_seq_length, num_encoder_tokens), dtype="float32"
)
decoder_input_data = np.zeros(
    (len(input_texts), max_decoder_seq_length, num_decoder_tokens), dtype="float32"
)
decoder_target_data = np.zeros(
    (len(input_texts), max_decoder_seq_length, num_decoder_tokens), dtype="float32"
)
#encoder输入是一句话，decoder输出也是一句话，即predict输出的outputs是一句话。
  
for i, (input_text, target_text) in enumerate(zip(input_texts, target_texts)):
    for t, char in enumerate(input_text):
        encoder_input_data[i, t, input_token_index[char]] = 1.0
    encoder_input_data[i, t + 1:, input_token_index[" "]] = 1.0 #句子补齐
    
    for t, char in enumerate(target_text):
        # decoder_target_data is ahead of decoder_input_data by one timestep
        decoder_input_data[i, t, target_token_index[char]] = 1.0
        if t > 0:
            # decoder_target_data will be ahead by one timestep
            # and will not include the start character.
            decoder_target_data[i, t - 1, target_token_index[char]] = 1.0
    decoder_input_data[i, t + 1:, target_token_index[" "]] = 1.0 #句子补齐
    decoder_target_data[i, t:, target_token_index[" "]] = 1.0  # 句子补齐
    

# build model
# Define an input sequence and process it.
encoder_inputs = keras.Input(shape=(None, num_encoder_tokens))
print('in',encoder_inputs.shape)
encoder = layers.LSTM(latent_dim,return_state=True)
encoder_outputs, state_h, state_c = encoder(encoder_inputs)

#这里如果return_sequences是true，那么输出的会是矩阵.默认是false,输出是一个向量hn，即只记录最后一个向量。后面会遇到attention模型，输出是一个矩阵，原因就是需要记住之前的所有状态h1,h2...hn，并和之前所有状态h1,h2...hn作对比。
print(encoder_outputs.shape)


# We discard `encoder_outputs` and only keep the states.
encoder_states = [state_h, state_c]

# Set up the decoder, using `encoder_states` as initial state.
decoder_inputs = keras.Input(shape=(None, num_decoder_tokens))

# We set up our decoder to return full output sequences,
# and to return internal states as well. We don't use the
# return states in the training model, but we will use them in inference.
decoder_lstm = layers.LSTM(
    latent_dim, return_sequences=True, return_state=True)
decoder_outputs, _, _ = decoder_lstm(
    decoder_inputs, initial_state=encoder_states)
print(decoder_outputs.shape)    
decoder_dense = layers.Dense(num_decoder_tokens, activation="softmax")
decoder_outputs = decoder_dense(decoder_outputs)

# Define the model that will turn
# `encoder_input_data` & `decoder_input_data` into `decoder_target_data`
model = keras.Model([encoder_inputs, decoder_inputs], decoder_outputs)


# train 
model.compile(
    optimizer="rmsprop", loss="categorical_crossentropy", metrics=["accuracy"]
)

model.fit(
    [encoder_input_data, decoder_input_data],
    decoder_target_data,
    batch_size=batch_size,
    epochs=epochs,
    validation_split=0.2,
)
# Save model
model.save("s2s")

(2)预测代码

model = keras.models.load_model("s2s")

model.summary()

encoder_inputs = model.input[0]  # input_1

encoder_outputs, state_h_enc, state_c_enc = model.layers[2].output  # lstm_1


encoder_states = [state_h_enc, state_c_enc]
#这个地方model的输出是encoder_states,与下面对应。decoder只需要接受一个来自encoder的状态就好。
encoder_model = keras.Model(encoder_inputs, encoder_states)

decoder_inputs = model.input[1]  # input_2

decoder_state_input_h = keras.Input(shape=(latent_dim,), name="input_3")
decoder_state_input_c = keras.Input(shape=(latent_dim,), name="input_4")

decoder_states_inputs = [decoder_state_input_h, decoder_state_input_c]
decoder_lstm = model.layers[3]

decoder_outputs, state_h_dec, state_c_dec = decoder_lstm(
    decoder_inputs, initial_state=decoder_states_inputs
)
decoder_states = [state_h_dec, state_c_dec]
decoder_dense = model.layers[4]

decoder_outputs = decoder_dense(decoder_outputs)

# print('decoder_outputs', decoder_outputs.shape)

decoder_model = keras.Model(
    [decoder_inputs] + decoder_states_inputs, [decoder_outputs] + decoder_states
)

# Reverse-lookup token index to decode sequences back to
# something readable.
reverse_input_char_index = dict((i, char)
                                for char, i in input_token_index.items())
reverse_target_char_index = dict((i, char)
                                 for char, i in target_token_index.items())


def decode_sequence(input_seq):
    # Encode the input as state vectors.
    states_value = encoder_model.predict(input_seq)

    # Generate empty target sequence of length 1.
    target_seq = np.zeros((1, 1, num_decoder_tokens))
    # Populate the first character of target sequence with the start character.
    target_seq[0, 0, target_token_index["\t"]] = 1.0
    
    # Sampling loop for a batch of sequences
    # (to simplify, here we assume a batch of size 1).
    stop_condition = False
    decoded_sentence = ""
    while not stop_condition:
        output_tokens, h, c = decoder_model.predict(
            [target_seq] + states_value)  # 此处的意思是将两个list（target_seq与states_value）合并
        # print('target seq',[target_seq])    
        # print('states value:',states_value)    
        # print('output_tokens', output_tokens.shape)
        # Sample a token
        sampled_token_index = np.argmax(output_tokens[0, -1, :])
        sampled_char = reverse_target_char_index[sampled_token_index]
        decoded_sentence += sampled_char

        # Exit condition: either hit max length
        # or find stop character.
        if sampled_char == "\n" or len(decoded_sentence) > max_decoder_seq_length:
            stop_condition = True

        # Update the target sequence (of length 1).
        target_seq = np.zeros((1, 1, num_decoder_tokens))
        target_seq[0, 0, sampled_token_index] = 1.0

        # just for test
        # target_seq = np.zeros((1, len(decoded_sentence), num_decoder_tokens))
        # for t, char in enumerate(decoded_sentence):
        #     target_seq[0,t,target_token_index[char]]=1.0    

        # Update states
        states_value = [h, c]
    return decoded_sentence

for seq_index in range(40):
    # Take one sequence (part of the training set)
    # for trying out decoding.
    input_seq = encoder_input_data[seq_index: seq_index + 1]
    decoded_sentence = decode_sequence(input_seq)
    print("-")
    print("Input sentence:", input_texts[seq_index])
    print("Decoded sentence:", decoded_sentence)