padding with max seq len

padding_example.py

import numpy as np
import tensorflow as tf

batch_size = 3
max_seq_len = 5
dim = 4

input_data = tf.placeholder(shape = (None,max_seq_len,dim), dtype = tf.float32)
seq_len = tf.placeholder(shape = [None], dtype = tf.int32)

max_len_range = np.array([range(max_seq_len)]*dim).T
max_len_range = np.reshape(max_len_range, (1,max_len_range.shape[0],max_len_range.shape[1]))

new_a = tf.tile(max_len_range, [batch_size,1,1])
seq_len_ = tf.reshape(seq_len, (-1,1,1))

max_len_mask = new_a < seq_len_
max_len_mask = tf.cast(max_len_mask , dtype = tf.float32)

pad_output = tf.multiply(input_data, max_len_mask)

sess = tf.Session()

sample_data  = np.random.rand(batch_size, max_seq_len, dim)
sample_seq_len = [3,4,2]

sess.run(pad_output, feed_dict = {input_data : sample_data, seq_len : sample_seq_len})