def word_dropout()

in access/preprocess.py [0:0]

• 6 lines of code
• 4 McCabe index (conditional complexity)

def word_dropout(words, dropout_prob=0.1):
    keep = torch.rand(len(words))
    dropped_out_words = [word for i, word in enumerate(words) if keep[i] > dropout_prob]
    if len(dropped_out_words) == 0:
        return [words[random.randint(0, len(words) - 1)]]
    return dropped_out_words