我看到很多样例代码中, 展开实现LSTM按时间步迭代时,都如下加上了一句
if timestep > 0:
tf.get_variable_scope().reuse_variables()
with tf.variable_scope('RNN'):
for timestep in range(timestep_size):
if timestep > 0:
tf.get_variable_scope().reuse_variables()
# 这里的state保存了每一层 LSTM 的状态
(cell_output, state) = mlstm_cell(X[:, timestep, :], state)
我查了各类RNNCELL类的源代码,发现除了call函数外,还有build函数(此函数似乎是在创建所有的变量),且此函数只调用一次,并且是在call调用之前调用,并且查看源码注释时发现是这样写的:
For backwards compatibility purposes, most RNNCell instances allow their call methods to instantiate variables via tf.get_variable. The underlying variable scope thus keeps track of any variables, and returning cached versions. This is atypical of tf.layer objects, which separate this
part of layer building into a build method that is only called once.
Here we provide a subclass for RNNCell objects that act exactly as
Layer objects do. They must provide a build method and their
call methods do not access Variables tf.get_variable
所以我个人认为这句不需要加,求各位帮忙解答下,该不该加这句话