LSTM

参数解析：

• init_scale – the initial scale of the weights
• learning_rate – the initial value of the learning rate
• max_grad_norm – the maximum permissible norm of the gradient
• num_layers – the number of LSTM layers
• num_steps – the number of unrolled steps of LSTM 这个指的就是time_step，也就是输入的词的个数
• hidden_size – the number of LSTM units 每一层lstm有多少个小单元
• max_epoch – the number of epochs trained with the initial learning rate
• max_max_epoch – the total number of epochs for training
• keep_prob – the probability of keeping weights in the dropout layer
• lr_decay – the decay of the learning rate for each epoch after “max_epoch”
• batch_size – the batch size

LSTM的输入

将embedding和input进行映射，使用embedding_lookup，每次输入的是[size_batches, seq_length, rnn_size]，三个参数分别是：时间长度，batch的size，rnn中的unit个数
之所以这样拆的原因是：
为使学习过程易于处理，通常的做法是将反向传播的梯度在（按时间）展开的步骤上照一个固定长度(seq_length)截断。 通过在一次迭代中的每个时刻上提供长度为 size_batch 的输入和每次迭代完成之后反向传导，这会很容易实现。

输入的变化：x_data = [446,50,50],指的是[number_batch,size_batch,seq_length] ==>embedding = [65,128],指的是[vocab_size,rnn_size] ==> Input_data=[50,50],指的是[batch_size,seq_length] ==> embedding_lookup(embedding,input_data) ==> [50,50,128],指的是[batch_size,seq_length,rnn_size]

LSTM初始化声明

lstm_cell = tf.nn.rnn_cell.BasicLSTMCell(size, forget_bias=0.0)
#size 指的就是hidden_size

lstm_cell = tf.nn.rnn_cell.DropoutWrapper(lstm_cell, output_keep_prob=config.keep_prob) # dropout的声明
cell = tf.nn.rnn_cell.MultiRNNCell([lstm_cell] * config.num_layers) #多层RNN的声明方式

LSTM的输入

用一个word2vec表示每个词语，输入的矩阵会被随机初始化，然后随着模型的学习，来不断修改

# embedding_matrix 张量的形状是： [vocabulary_size, embedding_size]
word_embeddings = tf.nn.embedding_lookup(embedding_matrix, word_ids)

LSTM训练过程

state = self._initial_state
    with tf.variable_scope("RNN"):
      for time_step in range(num_steps):
        if time_step > 0: tf.get_variable_scope().reuse_variables()
        (cell_output, state) = cell(inputs[:, time_step, :], state)
        outputs.append(cell_output)

LSTM误差声明

output = tf.reshape(tf.concat(1, outputs), [-1, size]) #
    softmax_w = tf.get_variable("softmax_w", [size, vocab_size])
    softmax_b = tf.get_variable("softmax_b", [vocab_size])
    logits = tf.matmul(output, softmax_w) + softmax_b
    loss = tf.nn.seq2seq.sequence_loss_by_example(
        [logits],
        [tf.reshape(self._targets, [-1])],
        [tf.ones([batch_size * num_steps])])
    self._cost = cost = tf.reduce_sum(loss) / batch_size

LSTM的迭代过程

for step, (x, y) in enumerate(reader.ptb_iterator(data, m.batch_size,
                                                    m.num_steps)):
    cost, state, _ = session.run([m.cost, m.final_state, eval_op],
                                 {m.input_data: x,
                                  m.targets: y,
                                  m.initial_state: state})
    costs += cost
    iters += m.num_steps

    if verbose and step % (epoch_size // 10) == 10:
      print("%.3f perplexity: %.3f speed: %.0f wps" %
            (step * 1.0 / epoch_size, np.exp(costs / iters),
             iters * m.batch_size / (time.time() - start_time)))

  return np.exp(costs / iters)
  # 此处需要针对cost，final_state，eval_op三个结构进行求解，输入三个参数如下，input_data，target，initial_state

返回的总误差是$$Loss = -\frac{1}{N}\sum_{i=1}^N InP_{target_i}$$
$$TotalLoss = e^{Loss}$$

  for i in range(config.max_max_epoch):
      lr_decay = config.lr_decay ** max(i - config.max_epoch, 0.0)
      m.assign_lr(session, config.learning_rate * lr_decay)

      print("Epoch: %d Learning rate: %.3f" % (i + 1, session.run(m.lr)))
      train_perplexity = run_epoch(session, m, train_data, m.train_op,
                                   verbose=True)

先完成model的初始化，然后在针对loss，train_op进行优化求解，通过SGD等方式进行求解

RNN-LSTM 参数设置

import argparse
parser = argparse.AugmentParser()
#添加参数名称，类型，缺省值，帮助提示
parser.add_argument('--batch_size',type=int,defaule = 50, help='mini batch size')
parser.add_argument('--learn_rate', type = float, default = 0.01, help = 'learn rate.')
parser.add_argument('--')

save and restore

model = Model(saved_args, True)
saver = tf.train.Saver(tf.all_variables())
with tf.Session() as sess:
        #tf.initialize_all_variables().run()

        sess.run(tf.initialize_all_variables())
        saver = tf.train.Saver(tf.all_variables())# save all variables 
        ckpt = tf.train.get_checkpoint_state(args.save_dir)
        if ckpt and ckpt.model_checkpoint_path:
            saver.restore(sess, ckpt.model_checkpoint_path)# restore the sess from ckpt.model_checkpoint_path
            print model.sample(sess, chars, vocab, args.n, args.prime)

tf.assign()