INNER CODE UNIT · Python
parse_file
HIT-SCIR/Chinese-Mixtral-8x7B · data/parquet2jsonl.py:9
def parse_file(filename):
pds = pq.read_pandas(filename, columns=None).to_pandas()
return pds.to_json(
path_or_buf=None,
orient="records",
lines=True,
date_format="iso",
date_unit="us",
compression="gzip",
).split("\n")
def main(
target_jsonl: str = "./data/en/DKYoon-SlimPajama-6B.jsonl",
root: str = "./data/en/DKYoon-SlimPajama-6B/data",
):
root = Path(root)
target_jsonl = Path(target_jsonl)