Datasets:
File size: 4,162 Bytes
5a2cdcc 165cabc 5a2cdcc 165cabc 5a2cdcc 8b73128 5a2cdcc 165cabc 5a2cdcc 8b73128 5a2cdcc 165cabc 5a2cdcc 165cabc 5a2cdcc 165cabc 5a2cdcc 165cabc | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 | """Download, verify and extract selected splits of the public release."""
from __future__ import annotations
import argparse
import json
import sys
import tarfile
from pathlib import Path
# GitHub: scripts/prepare_dataset.py; HF: prepare_dataset.py beside release_utils.py.
sys.path.insert(0, str(Path(__file__).resolve().parent.parent))
from release_utils import (ARTIFACTS, COUNTS, INDEX_SHA, PROTOCOL_SHA, REPO,
REVISION, SUMS_SHA, LICENSE_REVISION, LICENSE_FILES,
check_hash, contained_path, write_records)
def selected_archives(root, splits, sums):
selected = sorted(name for name in sums if name.endswith('.tar.gz')
and name.split('-', 1)[0] in splits)
if any(not any(name.startswith(s + '-') for name in selected) for s in splits):
raise ValueError('Checksum manifest is missing a requested split')
paths = [contained_path(root / 'archives', name) for name in selected]
for path in paths:
check_hash(path, sums[path.name])
return paths
def main():
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument('--local-dir', type=Path, default=Path('astrasr_data'))
parser.add_argument('--splits', nargs='+', choices=COUNTS, default=['val', 'test'])
parser.add_argument('--download-only', action='store_true')
parser.add_argument('--local-only', action='store_true', help='Verify/extract an existing download')
parser.add_argument('--artifacts', action='store_true', help='Also download weights and the PSF bank')
parser.add_argument('--remove-archives', action='store_true', help='Remove verified archives after extraction')
parser.add_argument('--keep-archives', action='store_true', help='Archives are kept by default')
args = parser.parse_args()
if args.remove_archives and args.keep_archives:
parser.error('--remove-archives conflicts with --keep-archives')
if not hasattr(tarfile, 'data_filter'):
parser.error('Use Python 3.12+ or a patched Python with tarfile.data_filter')
root = args.local_dir.resolve()
if not args.local_only:
from huggingface_hub import snapshot_download
patterns = ['SHA256SUMS.txt', 'dataset_index.jsonl', 'x2_dataset_protocol_lrdegrade_v2.json']
patterns += [f'archives/{split}-*.tar.gz' for split in args.splits]
if args.artifacts:
patterns += list(ARTIFACTS)
snapshot_download(REPO, repo_type='dataset', revision=REVISION,
local_dir=str(root), allow_patterns=patterns, max_workers=2)
snapshot_download(REPO, repo_type='dataset', revision=LICENSE_REVISION,
local_dir=str(root), allow_patterns=list(LICENSE_FILES), max_workers=2)
for name, digest in LICENSE_FILES.items():
check_hash(root / name, digest)
check_hash(root / 'SHA256SUMS.txt', SUMS_SHA)
check_hash(root / 'dataset_index.jsonl', INDEX_SHA)
check_hash(root / 'x2_dataset_protocol_lrdegrade_v2.json', PROTOCOL_SHA)
sums = {}
for line in (root / 'SHA256SUMS.txt').read_text(encoding='utf-8').splitlines():
digest, name = line.split(None, 1)
sums[name.strip()] = digest
archives = selected_archives(root, set(args.splits), sums)
if args.artifacts:
for name, digest in ARTIFACTS.items():
check_hash(root / name, digest)
print(f'Verified {len(archives)} requested archives at revision {REVISION}', flush=True)
if args.download_only:
return
output = root / 'extracted'
output.mkdir(parents=True, exist_ok=True)
for archive in archives:
print(f'Extracting {archive.name}', flush=True)
with tarfile.open(archive, 'r:gz') as stream:
stream.extractall(output, filter='data')
records = write_records(root, args.splits)
if args.remove_archives:
for archive in archives:
archive.unlink()
print(json.dumps({'records': str(records), 'splits': sorted(set(args.splits)),
'note': 'Local index paths differ from the historical server index.'}))
if __name__ == '__main__':
main()
|