WordPermutation

class opennmt.data.WordPermutation(max_distance)

Randomly permutes words in a sequence with a maximum distance.

Example

>>> noise = opennmt.data.WordPermutation(3)
>>> words = tf.constant(["0", "1", "2", "3", "4", "5", "6"])
>>> noise(words).numpy()
array([b'1', b'0', b'2', b'4', b'3', b'6', b'5'], dtype=object)

Inherits from: opennmt.data.Noise

__init__(max_distance)

Initializes the noise module.

Parameters

max_distance – The maximum permutation distance.