Dynamic augmentation methods in Locomotive

I’ve just added augmentation methods in Locomotive:

This allows Locomotive to add more samples from the training data. For example, the single_word_punctuation augmenter when passed:

('hello', 'salve')

Adds to the training data:

[('hello?', 'salve?'), ('hello!', 'salve!'), ('hello.', 'salve.')]

1 Like