scripts/build_hf_dataset.py

#!/usr/bin/env python3
"""Convert JSONL data files to HuggingFace-compatible parquet format."""

import subprocess
import sys
from pathlib import Path


def ensure_datasets():
    try:
        import datasets  # noqa: F401
    except ImportError:
        print("Installing datasets library...")
        subprocess.check_call([sys.executable, "-m", "pip", "install", "datasets"])


def main():
    ensure_datasets()
    from datasets import Dataset

    data_dir = Path(__file__).resolve().parent.parent / "data"
    hf_dir = Path(__file__).resolve().parent.parent / "hf"
    hf_dir.mkdir(exist_ok=True)

    configs = ["turns", "conversations", "sft_pairs", "chunks", "lexicon"]

    for name in configs:
        jsonl_path = data_dir / f"{name}.jsonl"
        if not jsonl_path.exists():
            print(f"SKIP {name}: {jsonl_path} not found")
            continue

        print(f"Loading {name}...")
        ds = Dataset.from_json(str(jsonl_path))
        out_path = hf_dir / f"{name}.parquet"
        ds.to_parquet(str(out_path))
        print(f"  -> {out_path} ({len(ds)} rows, {out_path.stat().st_size / 1e6:.1f} MB)")

    print("\nDone. Parquet files:")
    for f in sorted(hf_dir.glob("*.parquet")):
        print(f"  {f.name}: {f.stat().st_size / 1e6:.1f} MB")


if __name__ == "__main__":
    main()