How does rnn initialization behave?

Question

Python

I came across this code

# tf Graph input
x = tf.placeholder("float", [None, seq_max_len, 1])
y = tf.placeholder("float", [None, n_classes])
# A placeholder for indicating each sequence length
seqlen = tf.placeholder(tf.int32, [None])

# Define weights
weights = {
    'out': tf.Variable(tf.random_normal([n_hidden, n_classes]))
}
biases = {
    'out': tf.Variable(tf.random_normal([n_classes]))
}


def dynamicRNN(x, seqlen, weights, biases):

    # Prepare data shape to match `rnn` function requirements
    # Current data input shape: (batch_size, n_steps, n_input)
    # Required shape: 'n_steps' tensors list of shape (batch_size, n_input)
    
    # Unstack to get a list of 'n_steps' tensors of shape (batch_size, n_input)
    x = tf.unstack(x, seq_max_len, 1)

    # Define a lstm cell with tensorflow
    lstm_cell = tf.contrib.rnn.BasicLSTMCell(n_hidden)

    # Get lstm cell output, providing 'sequence_length' will perform dynamic
    # calculation.
    outputs, states = tf.contrib.rnn.static_rnn(lstm_cell, x, dtype=tf.float32,
                                sequence_length=seqlen)

    # When performing dynamic calculation, we must retrieve the last
    # dynamically computed output, i.e., if a sequence length is 10, we need
    # to retrieve the 10th output.
    # However TensorFlow doesn't support advanced indexing yet, so we build
    # a custom op that for each sample in batch size, get its length and
    # get the corresponding relevant output.

    # 'outputs' is a list of output at every timestep, we pack them in a Tensor
    # and change back dimension to [batch_size, n_step, n_input]
    outputs = tf.stack(outputs)
    outputs = tf.transpose(outputs, [1, 0, 2])

    # Hack to build the indexing and retrieve the right output.
    batch_size = tf.shape(outputs)[0]
    # Start indices for each sample
    index = tf.range(0, batch_size) * seq_max_len + (seqlen - 1)
    # Indexing
    outputs = tf.gather(tf.reshape(outputs, [-1, n_hidden]), index)

    # Linear activation, using outputs computed above
    return tf.matmul(outputs, weights['out']) + biases['out']

pred = dynamicRNN(x, seqlen, weights, biases)

# Define loss and optimizer
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pred, labels=y))
optimizer = tf.train.GradientDescentOptimizer(learning_rate=learning_rate).minimize(cost)

Actually a question:
Unless at each call to function dynamic_rnn lstm layers will not be re-initialized?

Answer 1

In TF, initialization is done via session.run(tf.global_variables_initializer())
Network weights "exist" only within the session. All work with them (initialization, modification, saving, loading) occurs through session.run