Buckets:
| """清理 bucket 内被取代的中间产物与日志。""" | |
| import os, shutil | |
| from pathlib import Path | |
| removed = [] | |
| def rm(p): | |
| p = Path(p) | |
| if p.is_dir(): | |
| sz = sum(f.stat().st_size for f in p.rglob('*') if f.is_file()) | |
| shutil.rmtree(p) | |
| removed.append((str(p), round(sz / 1e6, 1))) | |
| elif p.is_file(): | |
| sz = p.stat().st_size | |
| p.unlink() | |
| removed.append((str(p), round(sz / 1e6, 1))) | |
| # 1) 补齐版计数，已被 counts_ragged.npy 取代并上线 | |
| rm('/domains/counts-v3') | |
| # 2) 旧的 hierarchy.json（早已改名 labels.json，是遗留） | |
| rm('/domains/omnibrain-multi/hierarchy.json') | |
| # 3) 日志 | |
| for f in sorted(Path('/domains').glob('*.log')): | |
| rm(f) | |
| print('removed:') | |
| for p, sz in removed: | |
| print(f' {p} {sz} MB') | |
| print(f'total {sum(s for _, s in removed):.1f} MB') | |
| print() | |
| print('remaining top-level:') | |
| print(sorted(os.listdir('/domains'))) | |
Xet Storage Details
- Size:
- 915 Bytes
- Xet hash:
- 39e56f210047fd5490e6ea9097309e72b17db91201b937e5e129520947ce8075
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.