我知道有一些相关问题，但是我仍然无法解决我遇到的问题。我想训练LSTM用于情感分析，这是我的代码：
from sklearn.model_selection import train_test_split
train_x, test_x, train_y, test_y = train_test_split(dataset['reviewText'], dataset['label'], test_size = 0.1)
train_x = np.array(train_x)
train_y = np.array(train_y)
test_x = np.array(test_x)
test_y = np.array(test_y)
import tensorflow as tf
batch_size = 512
time_step = max_tokens
num_layers = 4
lstm_size = 256
learning_rate = 0.01
n_words = len(vocabulary)# Add 1 for 0 added to vocab
# Create the graph object
tf.reset_default_graph()
with tf.name_scope('inputs'):
inputs_ = tf.placeholder(tf.int32, [None, None], name="inputs")
labels_ = tf.placeholder(tf.int32, [None, None], name="labels")
keep_prob = tf.placeholder(tf.float64, name="keep_prob")
with tf.name_scope("Embeddings"):
embed = tf.nn.embedding_lookup(embedding_matrix, inputs_)
def lstm_cell():
lstm = tf.contrib.rnn.BasicLSTMCell(lstm_size, reuse=tf.get_variable_scope().reuse)
return tf.contrib.rnn.DropoutWrapper(lstm, output_keep_prob=keep_prob)
with tf.name_scope("RNN_layers"):
cell = tf.contrib.rnn.MultiRNNCell([lstm_cell() for _ in range(num_layers)])
initial_state = cell.zero_state(batch_size, tf.float64)
with tf.name_scope("RNN_forward"):
outputs, final_state = tf.nn.dynamic_rnn(cell, embed, initial_state=initial_state)
with tf.name_scope('predictions'):
predictions = tf.contrib.layers.fully_connected(outputs[:, -1], 1, activation_fn=tf.sigmoid)
tf.summary.histogram('predictions', predictions)
with tf.name_scope('cost'):
cost = tf.losses.mean_squared_error(labels_, predictions)
tf.summary.scalar('cost', cost)
with tf.name_scope('train'):
optimizer = tf.train.AdamOptimizer(learning_rate).minimize(cost)
merged = tf.summary.merge_all()
def get_batches(x, y, batch_size):
n_batches = len(x)//batch_size
x, y = x[:n_batches*batch_size], y[:n_batches*batch_size]
for ii in range(0, len(x), batch_size):
yield x[ii:ii+batch_size], y[ii:ii+batch_size]
epochs = 10
saver = tf.train.Saver()
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
train_writer = tf.summary.FileWriter('./logs/tb/train', sess.graph)
test_writer = tf.summary.FileWriter('./logs/tb/test', sess.graph)
iteration = 1
for e in range(epochs):
state = sess.run(initial_state)
for ii, (x, y) in enumerate(get_batches(train_x, train_y, batch_size), 1):
feed = {inputs_: x,
labels_: y[:, None],
keep_prob: 0.5,
initial_state: state}
print(type(merged))
print(type(cost))
print(type(final_state))
summary, loss, state, _ = sess.run([merged, cost, final_state, optimizer], feed_dict=feed)
train_writer.add_summary(summary, iteration)
if iteration%5==0:
print("Epoch: {}/{}".format(e, epochs),
"Iteration: {}".format(iteration),
"Train loss: {:.3f}".format(loss))
iteration +=1
test_writer.add_summary(summary, iteration)
saver.save(sess, "checkpoints/sentiment_manish.ckpt")
saver.save(sess, "checkpoints/sentiment_manish.ckpt")
问题是，每当我运行代码时，以下行：summary, loss, state, _ = sess.run([merged, cost, final_state, optimizer], feed_dict=feed)给出了我在问题标题中所述的错误。
编辑：
我在这里为您打印出这些变量的类型和值：
<class 'tensorflow.python.framework.ops.Tensor'>
Tensor("Merge/MergeSummary:0", shape=(), dtype=string)
<class 'tensorflow.python.framework.ops.Tensor'>
Tensor("cost/mean_squared_error/value:0", shape=(), dtype=float32)
<class 'tuple'>
(LSTMStateTuple(c=<tf.Tensor 'RNN_forward/rnn/while/Exit_3:0' shape=(512, 256) dtype=float64>, h=<tf.Tensor 'RNN_forward/rnn/while/Exit_4:0' shape=(512, 256) dtype=float64>), LSTMStateTuple(c=<tf.Tensor 'RNN_forward/rnn/while/Exit_5:0' shape=(512, 256) dtype=float64>, h=<tf.Tensor 'RNN_forward/rnn/while/Exit_6:0' shape=(512, 256) dtype=float64>), LSTMStateTuple(c=<tf.Tensor 'RNN_forward/rnn/while/Exit_7:0' shape=(512, 256) dtype=float64>, h=<tf.Tensor 'RNN_forward/rnn/while/Exit_8:0' shape=(512, 256) dtype=float64>), LSTMStateTuple(c=<tf.Tensor 'RNN_forward/rnn/while/Exit_9:0' shape=(512, 256) dtype=float64>, h=<tf.Tensor 'RNN_forward/rnn/while/Exit_10:0' shape=(512, 256) dtype=float64>))
我不确定我到底在做什么错，我认为final_state是一个元组存在问题。如果有人知道发生了什么事，我将不胜感激。
谢谢！