fairseq2plain.py

import argparse

"""
convert fairseq's output to plain text that have no extra information

"""

if __name__ == '__main__':

    parser = argparse.ArgumentParser()
    parser.add_argument("--input", type=str, required=True,
                        help="plain text")  # nargs="+",
    parser.add_argument("--output", type=str, required=True, )

    args = parser.parse_args()

    file_out = open(args.output, "wt", encoding="utf-8")

    with open(args.input, "rt", encoding='utf-8') as _file:
        for line in _file:
            seg_list = line.split('\t')
            if seg_list[0][0:2] == 'H-':
                file_out.write(seg_list[2])
                # file_out.write('\n')
    file_out.close()