INNER CODE UNIT · Python
extract_queries_split
UKPLab/gpl · gpl/toolkit/beir.py:9
def extract_queries_split(queries: Dict[str, str], qrels: Dict[str, Dict[str, float]]):
qids_to_exclude = set(qrels) - set(queries)
for qid in qids_to_exclude:
queries.pop(qid)
return queries
def save_queries(queries: Dict[str, str], output_dir):
os.makedirs(output_dir, exist_ok=True)
save_path = os.path.join(output_dir, "queries.jsonl")
with open(save_path, "w") as f:
for qid, query in queries.items():
line_dict = {"_id": qid, "text": query, "metadata": {}}
line = json.dumps(line_dict) + "\n"
f.write(line)
logger.info(f"Saved (copied) queries into {save_path}")