Example #1
0
if __name__ == '__main__':
    args = parser.parse_args()
    file_name = args.file_name
    dedup = args.dedup
    augment_size = args.augment_size
    seed = args.seed

    ner = Ner(ner_dir_name='.',
              ignore_tag_list=['O'],
              data_augument_tag_list=data_augument_tag_list,
              augment_size=augment_size,
              seed=seed,
              dedup=dedup)

    aug_samples, aug_sample_tags = ner.augment(file_name='%s.txt' % file_name)

    if dedup:
        target_dir = 'aug_dedup_%d' % augment_size
    else:
        target_dir = 'aug_%d' % augment_size
    if not os.path.exists(target_dir):
        os.makedirs(target_dir)
    with open(os.path.join(target_dir, '%s.txt' % file_name),
              'w',
              encoding='utf-8') as f:
        for tokens, tags in zip(aug_samples, aug_sample_tags):
            for token, tag in zip(tokens, tags):
                f.write("%s %s\n" % (token, tag))
            f.write("\n")