Dataset Viewer
Auto-converted to Parquet Duplicate
text
stringlengths
0
22
10:37-38
0:12
9:13
6:34
9:43
5:25
2:20
4:2
2:4
0:4
2:13
6:22
2:24
3:21
1:9 3:29
5:21
7:46
4:16
End of preview. Expand in Data Studio

compact-alignments-extra — opt-in residual layer

An optional, additive layer of compact-alignments. It holds ONLY the <BOOK>_<hash>.extra.json files, under the same relative path as in the main repo:

<iso[0]>/<iso>/<edition>/<BOOK>_<hash>.extra.json

For an alignment array at a/arb/arb_vdv/GEN_1a2b3.json in the main repo, this layer's file is a/arb/arb_vdv/GEN_1a2b3.extra.json here. The <hash> is the main file's content hash, so the two always belong together. Not every book has this layer; a missing file is a 404, not an error.

Content: the same compact string format as the main array, for target words a second, residual alignment pass could still explain after eflomal, gloss, span extension and gap-fill. It is deliberately NOT merged into the main array. Format: see the main repo's README (the r method letter).

Everything else — the README that defines the formats, manifest.json, _index/, tokenize.js, the decisions ledger — stays in the main repo. The authoritative list of editions and books is its manifest.json. The layer was split out because the single repo had passed Hugging Face's recommended 100,000 files.

License

CC0-1.0, like the main dataset. Each edition's own source and licence pointer is in the main repo's manifest.json.

Downloads last month
24