The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
words: struct<source: string, tag: string, file: string, sha256: string, licence: string, origin: string>
child 0, source: string
child 1, tag: string
child 2, file: string
child 3, sha256: string
child 4, licence: string
child 5, origin: string
layers: struct<hebrew-clause-layout: struct<tag: string, revision: string, licence: string, files: list<item (... 1755 chars omitted)
child 0, hebrew-clause-layout: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
child 1, greek-nt-clause-layout: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
child 2, clause-columns: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
child 3, clause-connectives: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
...
ring
child 1, tag: string
child 2, path: string
child 3, sha256: string
child 4, licence: string
child 5, origin: string
child 3, trees: struct<source: string, tag: string, path: string, sha256: string, licence: string, origin: string>
child 0, source: string
child 1, tag: string
child 2, path: string
child 3, sha256: string
child 4, licence: string
child 5, origin: string
dataset version: string
licence: string
code: string
schema: int64
queries: string
inputs: string
nodes: struct<text: int64, book: int64, person: int64, chapter: int64, lexeme: int64, quotation: int64, fac (... 144 chars omitted)
child 0, text: int64
child 1, book: int64
child 2, person: int64
child 3, chapter: int64
child 4, lexeme: int64
child 5, quotation: int64
child 6, fact: int64
child 7, verse: int64
child 8, sentence: int64
child 9, concept: int64
child 10, wordgroup: int64
child 11, clause: int64
child 12, line: int64
child 13, phrase: int64
child 14, token: int64
child 15, word: int64
built: timestamp[s]
version: string
flat tables: struct<lxx: int64, hebrew_nt: int64, quotes: int64>
child 0, lxx: int64
child 1, hebrew_nt: int64
child 2, quotes: int64
words per text: struct<tanakh: int64, greek_nt: int64, salkinson: int64, delitzsch: int64, septuagint: int64>
child 0, tanakh: int64
child 1, greek_nt: int64
child 2, salkinson: int64
child 3, delitzsch: int64
child 4, septuagint: int64
to
{'schema': Value('int64'), 'version': Value('string'), 'dataset version': Value('string'), 'built': Value('timestamp[s]'), 'code': Value('string'), 'licence': Value('string'), 'words per text': {'tanakh': Value('int64'), 'greek_nt': Value('int64'), 'salkinson': Value('int64'), 'delitzsch': Value('int64'), 'septuagint': Value('int64')}, 'nodes': {'text': Value('int64'), 'book': Value('int64'), 'person': Value('int64'), 'chapter': Value('int64'), 'lexeme': Value('int64'), 'quotation': Value('int64'), 'fact': Value('int64'), 'verse': Value('int64'), 'sentence': Value('int64'), 'concept': Value('int64'), 'wordgroup': Value('int64'), 'clause': Value('int64'), 'line': Value('int64'), 'phrase': Value('int64'), 'token': Value('int64'), 'word': Value('int64')}, 'flat tables': {'lxx': Value('int64'), 'hebrew_nt': Value('int64'), 'quotes': Value('int64')}, 'inputs': Value('string'), 'queries': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 558, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
words: struct<source: string, tag: string, file: string, sha256: string, licence: string, origin: string>
child 0, source: string
child 1, tag: string
child 2, file: string
child 3, sha256: string
child 4, licence: string
child 5, origin: string
layers: struct<hebrew-clause-layout: struct<tag: string, revision: string, licence: string, files: list<item (... 1755 chars omitted)
child 0, hebrew-clause-layout: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
child 1, greek-nt-clause-layout: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
child 2, clause-columns: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
child 3, clause-connectives: struct<tag: string, revision: string, licence: string, files: list<item: string>>
child 0, tag: string
child 1, revision: string
child 2, licence: string
child 3, files: list<item: string>
child 0, item: string
...
ring
child 1, tag: string
child 2, path: string
child 3, sha256: string
child 4, licence: string
child 5, origin: string
child 3, trees: struct<source: string, tag: string, path: string, sha256: string, licence: string, origin: string>
child 0, source: string
child 1, tag: string
child 2, path: string
child 3, sha256: string
child 4, licence: string
child 5, origin: string
dataset version: string
licence: string
code: string
schema: int64
queries: string
inputs: string
nodes: struct<text: int64, book: int64, person: int64, chapter: int64, lexeme: int64, quotation: int64, fac (... 144 chars omitted)
child 0, text: int64
child 1, book: int64
child 2, person: int64
child 3, chapter: int64
child 4, lexeme: int64
child 5, quotation: int64
child 6, fact: int64
child 7, verse: int64
child 8, sentence: int64
child 9, concept: int64
child 10, wordgroup: int64
child 11, clause: int64
child 12, line: int64
child 13, phrase: int64
child 14, token: int64
child 15, word: int64
built: timestamp[s]
version: string
flat tables: struct<lxx: int64, hebrew_nt: int64, quotes: int64>
child 0, lxx: int64
child 1, hebrew_nt: int64
child 2, quotes: int64
words per text: struct<tanakh: int64, greek_nt: int64, salkinson: int64, delitzsch: int64, septuagint: int64>
child 0, tanakh: int64
child 1, greek_nt: int64
child 2, salkinson: int64
child 3, delitzsch: int64
child 4, septuagint: int64
to
{'schema': Value('int64'), 'version': Value('string'), 'dataset version': Value('string'), 'built': Value('timestamp[s]'), 'code': Value('string'), 'licence': Value('string'), 'words per text': {'tanakh': Value('int64'), 'greek_nt': Value('int64'), 'salkinson': Value('int64'), 'delitzsch': Value('int64'), 'septuagint': Value('int64')}, 'nodes': {'text': Value('int64'), 'book': Value('int64'), 'person': Value('int64'), 'chapter': Value('int64'), 'lexeme': Value('int64'), 'quotation': Value('int64'), 'fact': Value('int64'), 'verse': Value('int64'), 'sentence': Value('int64'), 'concept': Value('int64'), 'wordgroup': Value('int64'), 'clause': Value('int64'), 'line': Value('int64'), 'phrase': Value('int64'), 'token': Value('int64'), 'word': Value('int64')}, 'flat tables': {'lxx': Value('int64'), 'hebrew_nt': Value('int64'), 'quotes': Value('int64')}, 'inputs': Value('string'), 'queries': Value('string')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Hebrew–Greek Bible (Context Fabric)
The Bible in Hebrew and Greek as one searchable Context Fabric dataset: the Tanakh, the Greek New Testament, the Septuagint and two Hebrew translations of the New Testament, on shared keys and linked word to word, with the bcv-commons analysis layers on the words. The links make it a bridge between Hebrew and Greek in both directions: from the Hebrew Bible to Greek through the Septuagint, and from the Greek New Testament to Hebrew through the Hebrew New Testaments.
| Version | v0.2 (dataset version 0.2+78a46690: the release and a hash of its pins) |
| Built | 2026-11-17 by macula/build_bridge_corpus.py and macula/release_hebrew_greek_cf.py (bcv-query 0417d0d) |
| Words | 1,450,126 in five texts |
| Licence | CC BY-SA 4.0 as a whole; every feature states its own source and licence |
| Checked | 28 fixed queries with known counts (queries.json) |
The texts
| Text | name |
Words | What it is |
|---|---|---|---|
| Tanakh | tanakh |
469,476 | MACULA Hebrew (Westminster Leningrad Codex), with the Aramaic; word parts as MACULA cuts them |
| Greek New Testament | greek_nt |
137,779 | MACULA Greek (Nestle 1904) |
| Hebrew New Testament, Salkinson-Ginsburg | salkinson |
179,037 | with bcv-commons/hebrew-nt-analysis |
| Hebrew New Testament, Delitzsch | delitzsch |
171,974 | with bcv-commons/hebrew-nt-analysis |
| Septuagint | septuagint |
491,860 | GLAUx's Septuagint (canonical books); Daniel both as the Old Greek and as Theodotion |
The base unit is word: a word part in Hebrew (as MACULA cuts it), a word in Greek. token is a written Hebrew word.
The links between the texts
| Edge | Pairs | Links |
|---|---|---|
lxx |
345,461 | Tanakh word part -> the Septuagint word that renders it; value: how it was placed (sure, order, neighbours, number) |
carried |
80,225 | Septuagint clause line -> the Tanakh clause line it carries over (a Hebrew line's Greek words, where they form a clean stretch) |
heb_scg |
123,728 | Greek NT word -> Salkinson-Ginsburg word part (hebrew-nt-bridge); value: the link's class |
heb_mod |
64,531 | Greek NT word -> Delitzsch word part (hebrew-nt-bridge); value: the link's class |
quotes |
1,298 | Hebrew NT verse -> the Tanakh verse it quotes; value: the weight |
parallel |
39,120 | verse -> the verse of the same number in the other text (Hebrew NT / Greek NT, Septuagint / Tanakh) |
Lexemes, concepts and persons are units shared by all five texts: a lexeme unit holds every occurrence of one Hebrew or Aramaic
lexeme (by Strong's number with MACULA's letter) or one Greek lemma, in any text; the tier_* edges link Hebrew and Greek lexemes
with their counts.
Units
| Unit | Count |
|---|---|
word |
1,450,126 |
token |
530,420 |
phrase |
203,505 |
line |
199,847 |
clause |
187,486 |
quotation |
6,782 |
verse |
70,364 |
chapter |
2,652 |
book |
160 |
text |
5 |
lexeme |
23,294 |
concept |
6,759 |
person |
536 |
fact |
62,663 |
sentence |
31,224 |
wordgroup |
442,131 |
Sections: sec on books (GEN, MAT; LXX.GEN, LXX.DAN-TH for Theodotion's Daniel; SCG.MAT, DEL.MAT for the Hebrew New
Testaments), then chapter and verse. book is the plain book name in every text.
The layers
Each layer is a published bcv-commons dataset at a pinned revision (pins.json). A layer's columns are features named
<prefix>_<column>.
| Layer | Tag | Revision | Prefix | Licence | Model-made |
|---|---|---|---|---|---|
| hebrew-clause-layout | v0.37.1 | 32f89b71 |
lay |
cc-by-sa-4.0 | no |
| greek-nt-clause-layout | v0.7 | 9168fac3 |
lay |
cc-by-4.0 | no |
| clause-columns | v1.1 | 4bf530de |
cc |
cc-by-sa-4.0 | no |
| clause-connectives | v0.1 | c26a84be |
cn |
cc-by-sa-4.0 | no |
| clause-extras | v0.1 | 5dbd5d06 |
cx |
cc-by-sa-4.0 | no |
| time-aspect | v0.2 | 5d30d2c2 |
ta |
cc-by-sa-4.0 | no |
| argument-roles | v0.2 | 5065acab |
ar |
cc-by-sa-4.0 | no |
| reference-status | v0.2 | 7612f663 |
rs |
cc-by-sa-4.0 | no |
| noun-relations | v0.1 | 73f86ec8 |
nr |
cc-by-sa-4.0 | no |
| word-senses | v0.1 | 00690ee3 |
ws |
cc-by-sa-4.0 | no |
| inclusive-we | v0.1 | d2d67cf4 |
we |
cc-by-sa-4.0 | no |
| addressees | v0.1 | 63683e3c |
ad |
cc-by-sa-4.0 | no |
| hebrew-word-concepts-model | v0.5 | 5a7af303 |
m |
cc-by-sa-4.0 | yes |
| hebrew-nt-analysis | v0.2 | 0c9132c3 |
hna |
cc-by-4.0 | no |
| hebrew-nt-bridge | v0.1 | 525db08f |
heb_scg, heb_mod, quotes, hnb |
cc-by-4.0 | no |
| word-parts | v0.1 | 1f32a49a |
wp |
cc-by-4.0 | no |
| hebrew-phrase-types-model | v0.1 | a5b12f05 |
m_pt |
cc-by-4.0 | yes |
| passage-facts | v0.1 | bbb9b2e3 |
pf |
cc-by-sa-4.0 | no |
Model-made layers (model_made=yes in the feature's metadata): hebrew-word-concepts-model, hebrew-phrase-types-model. A language model made these labels; no person has
checked them.
What is not in it: English content (glosses, English text columns), BHSA data, and anything from TBTA.
Loading
pip install context-fabric
from huggingface_hub import snapshot_download
import cfabric
path = snapshot_download("bcv-commons/hebrew-greek-cf", repo_type="dataset", revision="v0.2")
CF = cfabric.Fabric(locations=f"{path}/tf")
api = CF.loadAll()
The first load builds Context Fabric's cache beside the files (several minutes; 0.6: about 0.9 GB, 0.5: about 3.8 GB); later loads
take seconds. With Context Fabric 0.5.x, a reused cache forgets which edges carry values
(issue #15), so search refuses conditions like -lxx=sure>; set
the flag again after loading:
for name, feat in CF.features.items():
if getattr(getattr(api.E, name, None), "doValues", False):
feat.edgeValues = True
Context Fabric 0.6.0rc1 needs no such step (tested on this dataset).
A query, Hebrew word parts with a Septuagint word placed for sure:
word lang=hbo
-lxx=sure> word
Readers without Context Fabric: the links between the texts are also flat tables in bridges/ (Parquet), keyed by the words'
own ids (mid = the MACULA key, lkey = Septuagint word, hkey = Hebrew New Testament word part). In hebrew_nt, tanakh names
the Tanakh verse when the Hebrew word part lies in a stretch that follows that verse (a quotation the translator may have
taken from the Hebrew Bible rather than translated from the Greek); in the dataset this is the feature hnb_tanakh. Leave
these links out when counting the Hebrew NTs as evidence for a Greek-Hebrew pair.
Files
| Path | What it holds |
|---|---|
tf/ |
the dataset: one .tf file per feature |
bridges/*.parquet |
the links between texts as flat tables |
pins.json |
every input with its revision and licence |
report.json |
the build's counts |
queries.json |
the fixed queries and their counts |
build.json |
version, code, inputs |
Sources and licence
| Part | Source | Licence |
|---|---|---|
| Hebrew and Greek NT words, ids, lemmas, morphology | MACULA Hebrew (WLC) and MACULA Greek (Nestle 1904), Clear Bible | cc-by-4.0 |
| septuagint | GLAUx (KU Leuven), the Septuagint's canonical books; github:bcv-commons/spine-releases lxx-glaux-2026-10-10 | cc-by-sa-4.0 |
| theodotion | GLAUx (KU Leuven), Daniel by Theodotion (0527-057); github:bcv-commons/spine-releases lxx-glaux-dan-theodotion-2026-11-14 | cc-by-sa-4.0 |
| links | MACULA Hebrew's Septuagint forms (greek, greekstrong), placed by bridge_corpus_lxx; github:bcv-commons/spine-releases lexeme-spine-macula-2026-10-27 | cc-by-4.0 |
| trees | MACULA Hebrew (WLC lowfat) and MACULA Greek (Nestle 1904 lowfat) syntax trees, Clear Bible; github:bcv-commons/spine-releases trees-macula-2026-10-08 | cc-by-4.0 |
| the layers | bcv-commons datasets above | CC BY 4.0 / CC BY-SA 4.0 (per layer) |
The whole is CC BY-SA 4.0. Each feature's metadata names its source and licence, so that an output can state which features it read.
Versions
- v0.1: the first release (2026-11-17).
- v0.2:
hnb_tanakhon the Hebrew NT word parts andtanakhinbridges/hebrew_nt; the Septuagint tiers count both numbers of a compound Greek link (1722|1698,1909|848).
- Downloads last month
- 53