97 lines
4.4 KiB
Python
97 lines
4.4 KiB
Python
import tensorflow as tf
|
|
from tensorflow.keras.models import Sequential
|
|
from tensorflow.keras.layers import Embedding, LSTM, Dense
|
|
import numpy as np
|
|
|
|
# Загрузка и подготовка данных
|
|
with open('russian_text.txt', 'r', encoding='cp1251') as file:
|
|
russian_text = file.read()
|
|
|
|
# Токенизация текста
|
|
tokenizer = tf.keras.preprocessing.text.Tokenizer()
|
|
tokenizer.fit_on_texts([russian_text])
|
|
total_words = len(tokenizer.word_index) + 1
|
|
|
|
# Преобразование текста в числовые последовательности
|
|
input_sequences = []
|
|
for line in russian_text.split('n'):
|
|
token_list = tokenizer.texts_to_sequences([line])[0]
|
|
for i in range(1, len(token_list)):
|
|
n_gram_sequence = token_list[:i+1]
|
|
input_sequences.append(n_gram_sequence)
|
|
|
|
# Создание входных и целевых последовательностей
|
|
max_sequence_length = max([len(x) for x in input_sequences])
|
|
input_sequences = np.array(tf.keras.preprocessing.sequence.pad_sequences(input_sequences, maxlen=max_sequence_length,
|
|
padding='pre'))
|
|
predictors, label = input_sequences[:, :-1], input_sequences[:, -1]
|
|
|
|
# Построение и обучение модели
|
|
model = Sequential()
|
|
model.add(Embedding(total_words, 100, input_length=max_sequence_length-1))
|
|
model.add(LSTM(150))
|
|
model.add(Dense(total_words, activation='softmax'))
|
|
model.compile(loss='sparse_categorical_crossentropy', optimizer='adam')
|
|
model.fit(predictors, label, epochs=150, batch_size=64)
|
|
|
|
|
|
def generate_text(seed_text, next_words, model, max_sequence_length):
|
|
for _ in range(next_words):
|
|
token_list = tokenizer.texts_to_sequences([seed_text])[0]
|
|
token_list = tf.keras.preprocessing.sequence.pad_sequences([token_list], maxlen=max_sequence_length - 1,
|
|
padding='pre')
|
|
predicted = np.argmax(model.predict(token_list, verbose=0), axis=-1)
|
|
|
|
# Поиск слова, соответствующего предсказанному индексу
|
|
output_word = ""
|
|
for word, index in tokenizer.word_index.items():
|
|
if index == predicted:
|
|
output_word = word
|
|
break
|
|
|
|
seed_text += " " + output_word
|
|
|
|
return seed_text
|
|
|
|
# Генерация и сохранение русскоязычного текста
|
|
russian_generated_text = generate_text('Абрикосовая', next_words=100, model=model, max_sequence_length=max_sequence_length)
|
|
with open('generated_russian_text.txt', 'w', encoding='utf-8') as file:
|
|
file.write(russian_generated_text)
|
|
|
|
|
|
|
|
# Загрузка и подготовка данных
|
|
with open('english_text.txt', 'r', encoding='utf-8') as file:
|
|
english_text = file.read()
|
|
|
|
# Токенизация текста
|
|
tokenizer = tf.keras.preprocessing.text.Tokenizer()
|
|
tokenizer.fit_on_texts([english_text])
|
|
total_words = len(tokenizer.word_index) + 1
|
|
|
|
# Преобразование текста в числовые последовательности
|
|
input_sequences = []
|
|
for line in english_text.split('n'):
|
|
token_list = tokenizer.texts_to_sequences([line])[0]
|
|
for i in range(1, len(token_list)):
|
|
n_gram_sequence = token_list[:i+1]
|
|
input_sequences.append(n_gram_sequence)
|
|
|
|
# Создание входных и целевых последовательностей
|
|
max_sequence_length = max([len(x) for x in input_sequences])
|
|
input_sequences = np.array(tf.keras.preprocessing.sequence.pad_sequences(input_sequences, maxlen=max_sequence_length,
|
|
padding='pre'))
|
|
predictors, label = input_sequences[:, :-1], input_sequences[:, -1]
|
|
|
|
# Построение и обучение модели
|
|
model = Sequential()
|
|
model.add(Embedding(total_words, 100, input_length=max_sequence_length-1))
|
|
model.add(LSTM(150))
|
|
model.add(Dense(total_words, activation='softmax'))
|
|
model.compile(loss='sparse_categorical_crossentropy', optimizer='adam')
|
|
model.fit(predictors, label, epochs=150, batch_size=64)
|
|
|
|
# Генерация и сохранение англоязычного текста
|
|
english_generated_text = generate_text('It', next_words=100, model=model, max_sequence_length=max_sequence_length)
|
|
with open('generated_english_text.txt', 'w', encoding='utf-8') as file:
|
|
file.write(english_generated_text) |