model.add(Embedding(words, 10, input_length=input_len))
model.add(LSTM(100))
model.add(Dense(words, activation='softmax'))
model.compile(loss='categorical_crossentropy', optimizer='adam')
return model
model = create_model(max_len, words)
model.summary()
model.fit(predictors, label, epochs=50, verbose=2)
# Generating headlines based on input and LSTM
def headline_generator(text, next_words, model, max_len):
for _ in range(next_words):
token_list = tokenizer.texts_to_sequences([text])[0]
token_list = pad_sequences([token_list], maxlen=max_len-1, padding='pre')
predicted = model.predict_classes(token_list, verbose=0)
output_word = ""
for word,index in tokenizer.word_index.items():
if index == predicted:
output_word = word
