Buckets:

glennmatlin's picture
download
raw
1.45 kB
"""Generate manifests for enriched Dolma shards."""
from __future__ import annotations
import argparse
from pathlib import Path
from typing import Iterable
def build_parser() -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(description="Generate enriched shard manifest")
parser.add_argument("--input-dir", type=Path, default=Path("runs/dolma_enriched"))
parser.add_argument("--pattern", default="*.enriched.jsonl.zst")
parser.add_argument("--group-size", type=int, default=1)
parser.add_argument("--output", type=Path, required=True)
return parser
def _list_paths(input_dir: Path, pattern: str) -> list[Path]:
return sorted(input_dir.rglob(pattern))
def _chunk(items: list[Path], size: int) -> Iterable[list[Path]]:
for start in range(0, len(items), size):
yield items[start : start + size]
def main() -> int:
args = build_parser().parse_args()
if args.group_size <= 0:
raise ValueError("--group-size must be >= 1")
paths = _list_paths(args.input_dir, args.pattern)
if not paths:
raise ValueError("No enriched shards found for manifest")
lines = [
" ".join(str(path) for path in chunk)
for chunk in _chunk(paths, args.group_size)
]
args.output.parent.mkdir(parents=True, exist_ok=True)
args.output.write_text("\n".join(lines) + "\n", encoding="utf-8")
return 0
if __name__ == "__main__":
raise SystemExit(main())

Xet Storage Details

Size:
1.45 kB
·
Xet hash:
0c77266cabee5ef3f2023e4e37a596dab39c0ad5714e83032df6050936ad35a1

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.