How to use the tflearn.lstm function in tflearn

path = "US_Cities.txt"
if not os.path.isfile(path):
    context = ssl._create_unverified_context()
    moves.urllib.request.urlretrieve("https://raw.githubusercontent.com/tflearn/tflearn.github.io/master/resources/US_Cities.txt", path, context=context)

maxlen = 20

string_utf8 = open(path, "r").read().decode('utf-8')
X, Y, char_idx = \
    string_to_semi_redundant_sequences(string_utf8, seq_maxlen=maxlen, redun_step=3)

g = tflearn.input_data(shape=[None, maxlen, len(char_idx)])
g = tflearn.lstm(g, 512, return_seq=True)
g = tflearn.dropout(g, 0.5)
g = tflearn.lstm(g, 512)
g = tflearn.dropout(g, 0.5)
g = tflearn.fully_connected(g, len(char_idx), activation='softmax')
g = tflearn.regression(g, optimizer='adam', loss='categorical_crossentropy',
                       learning_rate=0.001)

m = tflearn.SequenceGenerator(g, dictionary=char_idx,
                              seq_maxlen=maxlen,
                              clip_gradients=5.0,
                              checkpoint_path='model_us_cities')

for i in range(40):
    seed = random_sequence_from_string(string_utf8, maxlen)
    m.fit(X, Y, validation_set=0.1, batch_size=128,
          n_epoch=1, run_id='us_cities')
    print("-- TESTING...")
    print("-- Test with temperature of 1.2 --")

valid_portion=0.1)
trainX, trainY = train
testX, testY = test

# Data preprocessing
# Sequence padding
trainX = pad_sequences(trainX, maxlen=100, value=0.)
testX = pad_sequences(testX, maxlen=100, value=0.)
# Converting labels to binary vectors
trainY = to_categorical(trainY)
testY = to_categorical(testY)

# Network building
net = tflearn.input_data([None, 100])
net = tflearn.embedding(net, input_dim=10000, output_dim=128)
net = tflearn.lstm(net, 128, dropout=0.8)
net = tflearn.fully_connected(net, 2, activation='softmax')
net = tflearn.regression(net, optimizer='adam', learning_rate=0.001,
                         loss='categorical_crossentropy')

# Training
model = tflearn.DNN(net, tensorboard_verbose=0)
model.fit(trainX, trainY, validation_set=(testX, testY), show_metric=True,
          batch_size=32)

# 预测过程用前一个时间序的输出作为下一个时间序的输入
        # 先用编码器的最后一个输出作为第一个输入
        if feed_previous:
            first_dec_input = go_inputs
        else:
            first_dec_input = tf.slice(decoder_inputs, [0, 0, 0], [-1, 1, self.word_vec_dim])
        decoder_output_tensor = tflearn.lstm(first_dec_input, self.word_vec_dim, initial_state=states, return_seq=False, reuse=False, scope='decoder_lstm')
        decoder_output_sequence_single = tf.pack([decoder_output_tensor], axis=1)
        decoder_output_sequence_list = [decoder_output_tensor]
        # 再用解码器的输出作为下一个时序的输入
        for i in range(self.max_seq_len-1):
            if feed_previous:
                next_dec_input = decoder_output_sequence_single
            else:
                next_dec_input = tf.slice(decoder_inputs, [0, i+1, 0], [-1, 1, self.word_vec_dim])
            decoder_output_tensor = tflearn.lstm(next_dec_input, self.word_vec_dim, return_seq=False, reuse=True, scope='decoder_lstm')
            decoder_output_sequence_single = tf.pack([decoder_output_tensor], axis=1)
            decoder_output_sequence_list.append(decoder_output_tensor)

        decoder_output_sequence = tf.pack(decoder_output_sequence_list, axis=1)
        real_output_sequence = tf.concat(1, [encoder_output_sequence, decoder_output_sequence])

        net = tflearn.regression(real_output_sequence, optimizer='sgd', learning_rate=0.1, loss='mean_square')
        model = tflearn.DNN(net)
        return model

def model(self, feed_previous=False):
        # 通过输入的XY生成encoder_inputs和带GO头的decoder_inputs
        input_data = tflearn.input_data(shape=[None, self.max_seq_len*2, self.word_vec_dim], dtype=tf.float32, name = "XY")
        encoder_inputs = tf.slice(input_data, [0, 0, 0], [-1, self.max_seq_len, self.word_vec_dim], name="enc_in")
        decoder_inputs_tmp = tf.slice(input_data, [0, self.max_seq_len, 0], [-1, self.max_seq_len-1, self.word_vec_dim], name="dec_in_tmp")
        go_inputs = tf.ones_like(decoder_inputs_tmp)
        go_inputs = tf.slice(go_inputs, [0, 0, 0], [-1, 1, self.word_vec_dim])
        decoder_inputs = tf.concat(1, [go_inputs, decoder_inputs_tmp], name="dec_in")

        # 编码器
        # 把encoder_inputs交给编码器，返回一个输出(预测序列的第一个值)和一个状态(传给解码器)
        (encoder_output_tensor, states) = tflearn.lstm(encoder_inputs, self.word_vec_dim, return_state=True, scope='encoder_lstm')
        encoder_output_sequence = tf.pack([encoder_output_tensor], axis=1)

        # 解码器
        # 预测过程用前一个时间序的输出作为下一个时间序的输入
        # 先用编码器的最后一个输出作为第一个输入
        if feed_previous:
            first_dec_input = go_inputs
        else:
            first_dec_input = tf.slice(decoder_inputs, [0, 0, 0], [-1, 1, self.word_vec_dim])
        decoder_output_tensor = tflearn.lstm(first_dec_input, self.word_vec_dim, initial_state=states, return_seq=False, reuse=False, scope='decoder_lstm')
        decoder_output_sequence_single = tf.pack([decoder_output_tensor], axis=1)
        decoder_output_sequence_list = [decoder_output_tensor]
        # 再用解码器的输出作为下一个时序的输入
        for i in range(self.max_seq_len-1):
            if feed_previous:
                next_dec_input = decoder_output_sequence_single

def make_network(look_back, batch_size):
    """
    Declare the layer types and sizes
    """
    # create deep neural network with LSTM and fully connected layers
    net = tfl.input_data(shape=[None, look_back, 1], name='input')
    net = tfl.lstm(net, 32, activation='tanh', weights_init='xavier', name='LSTM1')

    net = tfl.fully_connected(net, 20, activation='relu', name='FC1')
    # net = tfl.dropout(net, 0.5)
    net = tfl.fully_connected(net, 40, activation='relu', name='FC2')
    # net = tfl.dropout(net, 0.5)

    net = tfl.fully_connected(net, 1, activation='linear', name='Linear')
    net = tfl.regression(net, batch_size=batch_size, optimizer='adam', learning_rate=0.005, loss='mean_square', name='target')
    col = tf.get_collection(tf.GraphKeys.TRAINABLE_VARIABLES)
    for x in col:
        tf.add_to_collection(tf.GraphKeys.VARIABLES, x ) 
    return net

# 预测过程用前一个时间序的输出作为下一个时间序的输入
        # 先用编码器的最后一个输出作为第一个输入
        if feed_previous:
            first_dec_input = go_inputs
        else:
            first_dec_input = tf.slice(decoder_inputs, [0, 0, 0], [-1, 1, self.word_vec_dim])
        decoder_output_tensor = tflearn.lstm(first_dec_input, self.word_vec_dim, initial_state=states, return_seq=False, reuse=False, scope='decoder_lstm')
        decoder_output_sequence_single = tf.pack([decoder_output_tensor], axis=1)
        decoder_output_sequence_list = [decoder_output_tensor]
        # 再用解码器的输出作为下一个时序的输入
        for i in range(self.max_seq_len-1):
            if feed_previous:
                next_dec_input = decoder_output_sequence_single
            else:
                next_dec_input = tf.slice(decoder_inputs, [0, i+1, 0], [-1, 1, self.word_vec_dim])
            decoder_output_tensor = tflearn.lstm(next_dec_input, self.word_vec_dim, return_seq=False, reuse=True, scope='decoder_lstm')
            decoder_output_sequence_single = tf.pack([decoder_output_tensor], axis=1)
            decoder_output_sequence_list.append(decoder_output_tensor)

        decoder_output_sequence = tf.pack(decoder_output_sequence_list, axis=1)
        real_output_sequence = tf.concat(1, [encoder_output_sequence, decoder_output_sequence])

        net = tflearn.regression(real_output_sequence, optimizer='sgd', learning_rate=0.1, loss='mean_square')
        model = tflearn.DNN(net)
        return model

def get_network(frames, input_size, num_classes):
    """Create our LSTM"""
    net = tflearn.input_data(shape=[None, frames, input_size])
    net = tflearn.lstm(net, 128, dropout=0.8, return_seq=True)
    net = tflearn.lstm(net, 128)
    net = tflearn.fully_connected(net, num_classes, activation='softmax')
    net = tflearn.regression(net, optimizer='adam',
                             loss='categorical_crossentropy', name="output1")
    return net

def build_model(maxlen, char_idx, checkpoint_path):
    g = tflearn.input_data([None, maxlen, len(char_idx)])
    g = tflearn.lstm(g, 512, return_seq=True)
    g = tflearn.dropout(g, 0.5)
    g = tflearn.lstm(g, 512, return_seq=True)
    g = tflearn.dropout(g, 0.5)
    g = tflearn.lstm(g, 512)
    g = tflearn.dropout(g, 0.5)
    g = tflearn.fully_connected(g, len(char_idx), activation='softmax')
    g = tflearn.regression(g, optimizer='adam', loss='categorical_crossentropy',
                           learning_rate=0.001)

    return tflearn.SequenceGenerator(g, dictionary=char_idx,
                                     seq_maxlen=maxlen,
                                     clip_gradients=5.0,
                                     checkpoint_path=checkpoint_path)

def lstm_block(input, hidden_units=128, dropout=0.5, reuse=False, layers=1,
                           dynamic=True, return_seq=False, bidirectional=False):
    output = None
    prev_output = input
    for n_layer in range(layers):
        if not bidirectional:
            if n_layer < layers - 1:
                output = tflearn.lstm(prev_output, hidden_units, dropout=dropout,
                                dynamic=dynamic, reuse=reuse,
                                scope='lstm_{}'.format(n_layer), return_seq=True)
                output = tf.stack(output, axis=0)
                output = tf.transpose(output, perm=[1, 0, 2])
                prev_output = output
                continue
            output = tflearn.lstm(prev_output, hidden_units, dropout=dropout,
                                  dynamic=dynamic, reuse=reuse,
                                  scope='lstm_{}'.format(n_layer),
                                  return_seq=return_seq)
        else:
            if n_layer < layers - 1:
                output = bidirectional_rnn(prev_output,
                                           BasicLSTMCell(hidden_units,
                                                         reuse=reuse),
                                           BasicLSTMCell(hidden_units,