不能用Python TensorFlow中的LSTMBlockFusedCell替换LSTMBlockCell

Question

用LSTMBlockCell替换LSTMBlockFusedCell会在static_rnn`中引发TypeError。我正在使用从源代码编译的TensorFlow 1.2.0-rc1。

完整的错误消息：

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-3-2986e054cb6b> in <module>()
     19     enc_cell = tf.contrib.rnn.LSTMBlockFusedCell(rnn_size)
     20     enc_layers = tf.contrib.rnn.MultiRNNCell([enc_cell] * num_layers, state_is_tuple=True)
---> 21     _, enc_state = tf.contrib.rnn.static_rnn(enc_layers, enc_input_unstacked, dtype=dtype)
     22 
     23 with tf.variable_scope('decoder'):

~/Virtualenvs/scikit/lib/python3.6/site-packages/tensorflow/python/ops/rnn.py in static_rnn(cell, inputs, initial_state, dtype, sequence_length, scope)
   1139 
   1140   if not _like_rnncell(cell):
-> 1141     raise TypeError("cell must be an instance of RNNCell")
   1142   if not nest.is_sequence(inputs):
   1143     raise TypeError("inputs must be a sequence")

TypeError: cell must be an instance of RNNCell

代码重现：

import tensorflow as tf

batch_size = 8
enc_input_length = 1000

dtype = tf.float32
rnn_size = 8
num_layers = 2

enc_input = tf.placeholder(dtype, shape=[batch_size, enc_input_length, 1])
enc_input_unstacked = tf.unstack(enc_input, axis=1)

with tf.variable_scope('encoder'):
    enc_cell = tf.contrib.rnn.LSTMBlockFusedCell(rnn_size)
    enc_layers = tf.contrib.rnn.MultiRNNCell([enc_cell] * num_layers)
    _, enc_state = tf.contrib.rnn.static_rnn(enc_layers, enc_input_unstacked, dtype=dtype)

_like_rnncell看起来像：

def _like_rnncell(cell):
  """Checks that a given object is an RNNCell by using duck typing."""
  conditions = [hasattr(cell, "output_size"), hasattr(cell, "state_size"),
                hasattr(cell, "zero_state"), callable(cell)]
  return all(conditions)

原来LSTMBlockFusedCell没有output_size实施的state_size和LSTMBlockCell属性。

这是一个错误，还是有办法使用我缺少的LSTMBlockFusedCell。

Answer 1

LSTMBlockFusedCell继承自FusedRNNCell而不是RNNCell，所以你不能使用标准的tf.nn.static_rnn或tf.nn.dynamic_rnn，它们需要RNNCell实例（如你的错误信息所示）。

但是，在documentation中，您可以直接调用单元格以获得完整的输出和状态。

inputs = tf.placeholder(tf.float32, [time_len, batch_size, input_size])
fused_rnn_cell = tf.contrib.rnn.LSTMBlockFusedCell(num_units)

outputs, state = fused_rnn_cell(inputs, dtype=tf.float32)

# outputs shape is (time_len, batch_size, num_units)
# state: LSTMStateTuple where c shape is (batch_size, num_units)
#  and h shape is also (batch_size, num_units).

LSTMBlockFusedCell对象调用gen_lstm_ops.block_lstm internally，它应该等同于正常的LSTM循环。

另外，请注意任何FusedRNNCell实例的输入应该是时间主要的，这可以通过在调用单元格之前转置张量来完成。