def rnn(input_data, weights, biases):
# hidden layer for input to cell
########################################
# transpose the inputs shape from
# X?128 batch ,28 steps, 18 inputs?
# ==> (128 batch * 28 steps, 28 inputs)
input_data = tf.reshape(input_data, [-1, n_inputs])
# into hidden
# data_in = (128 batch * 28 steps, 128 hidden)
data_in = tf.matmul(input_data, weights['in']) + biases['in']
# data_in ==> (128 batch, 28 steps, 128 hidden_units)
data_in = tf.reshape(data_in, [-1, n_steps, n_hidden_units])
# cell
##########################################
# basic LSTM Cell.
lstm_cell = tf.nn.rnn_cell.BasicLSTMCell(n_hidden_units, forget_bias=1.0, state_is_tuple=True)
# lstm cell is divided into two parts (c_state, h_state)
_init_state = lstm_cell.zero_state(batch_size, dtype=tf.float32)
# You have 2 options for following step.
# 1: tf.nn.rnn(cell, inputs);
# 2: tf.nn.dynamic_rnn(cell, inputs).
# If use option 1, you have to modified the shape of data_in, go and check out this:
# https://github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/3_NeuralNetworks/recurrent_network.py
# In here, we go for option 2.
# dynamic_rnn receive Tensor (batch, steps, inputs) or (steps, batch, inputs) as data_in.
# Make sure the time_major is changed accordingly.
outputs, final_state = tf.nn.dynamic_rnn(lstm_cell, data_in, initial_state=_init_state, time_major=False)
# hidden layer for output as the final results
#############################################
# results = tf.matmul(final_state[1], weights_1['out']) + biases_1['out']
# # or
# unpack to list [(batch, outputs)..] * steps
outputs = tf.unpack(tf.transpose(outputs, [1, 0, 2])) # states is the last outputs
results = tf.matmul(outputs[-1], weights['out']) + biases['out']
return results
评论列表
文章目录
