| 13 | # 1. CLI PARSER |
| 14 | # --------------------------------------------------------------------------- # |
| 15 | def parse_args(): |
| 16 | p = argparse.ArgumentParser( |
| 17 | description="Process and split classification JSONL data" |
| 18 | ) |
| 19 | p.add_argument("--input_file", required=True, help="Path to the input JSONL file") |
| 20 | p.add_argument("--output_file", required=True, help="Path to the output JSONL file") |
| 21 | p.add_argument( |
| 22 | "--do_plot", action="store_true", help="Plot distributions of CCP and max_prob" |
| 23 | ) |
| 24 | p.add_argument( |
| 25 | "--all_info", |
| 26 | action="store_true", |
| 27 | help="Treat *all* primary_tag=='information seeking' rows as info-seeking " |
| 28 | "(ignore other_tags)", |
| 29 | ) |
| 30 | p.add_argument( |
| 31 | "--only_info", |
| 32 | action="store_true", |
| 33 | help="After processing, keep only info-seeking rows in the final output", |
| 34 | ) |
| 35 | p.add_argument( |
| 36 | "--no_surgery", |
| 37 | action="store_true", |
| 38 | help="Disable surgery / reflection logic (outputs plain responses)", |
| 39 | ) |
| 40 | return p.parse_args() |
| 41 | |
| 42 | |
| 43 | args = parse_args() |