from tensorflow.keras.models import Sequential

from tensorflow.keras.layers import Embedding, LSTM, Dense

def create_rnn_model(input_shape, num_classes):

    model = Sequential()

    model.add(Embedding(input_dim=10000, output_dim=128, input_length=input_shape[1]))

    model.add(LSTM(64))

    model.add(Dense(128, activation='relu'))

    model.add(Dense(num_classes, activation='softmax'))

    return model