Document storage requirements
Browse files
README.md
CHANGED
|
@@ -20,6 +20,24 @@ The original directory contains 1,841,072 files across:
|
|
| 20 |
To keep the Hugging Face dataset repository usable, the data is stored as split
|
| 21 |
`.tar.zst` archives instead of uploading the raw file tree.
|
| 22 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
## Download
|
| 24 |
|
| 25 |
Install the Hugging Face CLI if needed:
|
|
|
|
| 20 |
To keep the Hugging Face dataset repository usable, the data is stored as split
|
| 21 |
`.tar.zst` archives instead of uploading the raw file tree.
|
| 22 |
|
| 23 |
+
## Storage Requirements
|
| 24 |
+
|
| 25 |
+
The uploaded split archives take about `119G` on disk (`126.9GB` decimal).
|
| 26 |
+
After extraction, the restored `spatialcode_data/` directory takes about `127G`
|
| 27 |
+
on disk (`130.9GB` decimal).
|
| 28 |
+
|
| 29 |
+
Plan for:
|
| 30 |
+
|
| 31 |
+
- `119G` to download and store only the compressed archive parts.
|
| 32 |
+
- `127G` for the fully extracted `spatialcode_data/` directory.
|
| 33 |
+
- About `246G` total if you keep both the downloaded archives and extracted
|
| 34 |
+
data at the same time.
|
| 35 |
+
- `270G+` free space is recommended to leave room for filesystem overhead and
|
| 36 |
+
partial downloads.
|
| 37 |
+
|
| 38 |
+
If disk space is tight, restore one source at a time and delete its archive
|
| 39 |
+
parts after verifying extraction.
|
| 40 |
+
|
| 41 |
## Download
|
| 42 |
|
| 43 |
Install the Hugging Face CLI if needed:
|