Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: TypeError
Message: Couldn't cast array of type
struct<adv>adv: struct<kept: int64, rate: double, total: int64>, adv>o: struct<kept: int64, rate: double, total: int64>, adv>p: struct<kept: int64, rate: double, total: int64>, adv>pp: struct<kept: int64, rate: double, total: int64>, adv>s: struct<kept: int64, rate: double, total: int64>, adv>v: struct<kept: int64, rate: double, total: int64>, o2>o: struct<kept: int64, rate: double, total: int64>, o2>pp: struct<kept: int64, rate: double, total: int64>, o2>s: struct<kept: int64, rate: double, total: int64>, o2>v: struct<kept: int64, rate: double, total: int64>, o>adv: struct<kept: int64, rate: double, total: int64>, o>o: struct<kept: int64, rate: double, total: int64>, o>o2: struct<kept: int64, rate: double, total: int64>, o>p: struct<kept: int64, rate: double, total: int64>, o>pp: struct<kept: int64, rate: double, total: int64>, o>s: struct<kept: int64, rate: double, total: int64>, o>v: struct<kept: int64, rate: double, total: int64>, p>adv: struct<kept: int64, rate: double, total: int64>, p>o: struct<kept: int64, rate: double, total: int64>, p>p: struct<kept: int64, rate: double, total: int64>, p>pp: struct<kept: int64, rate: double, total: int64>, p>s: struct<kept: int64, rate: double, total: int64>, p>v: struct<kept: int64, rate: double, total: int64>, pp>adv: struct<kept: int64, rate: double, total: int64>, pp>o: struct<kept: int64, rate: double, total: int64>, pp>o2: struct<kept: int64, rate: double, total: int64>, pp>p: struct<kept: int64, rate: double, total: int64>, pp>pp: struct<kept: int64, rate: double, total: int64>, pp>s: struct<kept: int64, rate: double, total: int64>, pp>v: struct<kept: int64, rate: double, total: int64>, s>adv: struct<kept: int64, rate: double, total: int64>, s>o: struct<kept: int64, rate: double, total: int64>, s>o2: struct<kept: int64, rate: double, total: int64>, s>p: struct<kept: int64, rate: double, total: int64>, s>pp: struct<kept: int64, rate: double, total: int64>, s>s: struct<kept: int64, rate: double, total: int64>, s>v: struct<kept: int64, rate: double, total: int64>, v>adv: struct<kept: int64, rate: double, total: int64>, v>o: struct<kept: int64, rate: double, total: int64>, v>o2: struct<kept: int64, rate: double, total: int64>, v>p: struct<kept: int64, rate: double, total: int64>, v>pp: struct<kept: int64, rate: double, total: int64>, v>s: struct<kept: int64, rate: double, total: int64>, v>v: struct<kept: int64, rate: double, total: int64>>
to
{'adv>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'adv>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>o2': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'p>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'p>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'p>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>p': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>adv': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>p': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>adv': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>p': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}}
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 147, 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 127, 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 483, 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 2951, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2461, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2486, 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 547, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 430, 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 343, 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 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2312, in cast_table_to_schema
cast_array_to_feature(
~~~~~~~~~~~~~~~~~~~~~^
table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
feature,
^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1861, in wrapper
return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
~~~~^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2158, in cast_array_to_feature
raise TypeError(f"Couldn't cast array of type\n{_short_str(array.type)}\nto\n{_short_str(feature)}")
TypeError: Couldn't cast array of type
struct<adv>adv: struct<kept: int64, rate: double, total: int64>, adv>o: struct<kept: int64, rate: double, total: int64>, adv>p: struct<kept: int64, rate: double, total: int64>, adv>pp: struct<kept: int64, rate: double, total: int64>, adv>s: struct<kept: int64, rate: double, total: int64>, adv>v: struct<kept: int64, rate: double, total: int64>, o2>o: struct<kept: int64, rate: double, total: int64>, o2>pp: struct<kept: int64, rate: double, total: int64>, o2>s: struct<kept: int64, rate: double, total: int64>, o2>v: struct<kept: int64, rate: double, total: int64>, o>adv: struct<kept: int64, rate: double, total: int64>, o>o: struct<kept: int64, rate: double, total: int64>, o>o2: struct<kept: int64, rate: double, total: int64>, o>p: struct<kept: int64, rate: double, total: int64>, o>pp: struct<kept: int64, rate: double, total: int64>, o>s: struct<kept: int64, rate: double, total: int64>, o>v: struct<kept: int64, rate: double, total: int64>, p>adv: struct<kept: int64, rate: double, total: int64>, p>o: struct<kept: int64, rate: double, total: int64>, p>p: struct<kept: int64, rate: double, total: int64>, p>pp: struct<kept: int64, rate: double, total: int64>, p>s: struct<kept: int64, rate: double, total: int64>, p>v: struct<kept: int64, rate: double, total: int64>, pp>adv: struct<kept: int64, rate: double, total: int64>, pp>o: struct<kept: int64, rate: double, total: int64>, pp>o2: struct<kept: int64, rate: double, total: int64>, pp>p: struct<kept: int64, rate: double, total: int64>, pp>pp: struct<kept: int64, rate: double, total: int64>, pp>s: struct<kept: int64, rate: double, total: int64>, pp>v: struct<kept: int64, rate: double, total: int64>, s>adv: struct<kept: int64, rate: double, total: int64>, s>o: struct<kept: int64, rate: double, total: int64>, s>o2: struct<kept: int64, rate: double, total: int64>, s>p: struct<kept: int64, rate: double, total: int64>, s>pp: struct<kept: int64, rate: double, total: int64>, s>s: struct<kept: int64, rate: double, total: int64>, s>v: struct<kept: int64, rate: double, total: int64>, v>adv: struct<kept: int64, rate: double, total: int64>, v>o: struct<kept: int64, rate: double, total: int64>, v>o2: struct<kept: int64, rate: double, total: int64>, v>p: struct<kept: int64, rate: double, total: int64>, v>pp: struct<kept: int64, rate: double, total: int64>, v>s: struct<kept: int64, rate: double, total: int64>, v>v: struct<kept: int64, rate: double, total: int64>>
to
{'adv>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'adv>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>o2': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'o>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'p>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'p>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'p>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>p': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'pp>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>adv': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>p': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 's>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>adv': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>o': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>p': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>pp': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>s': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}, 'v>v': {'kept': Value('int64'), 'rate': Value('float64'), 'total': Value('int64')}}Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
constituent-order-profile
Per-language constituent-order statistics — how a language reorders Hebrew's clause-level syntax
(the seven MACULA Hebrew phrase_role labels: v verb, s subject, o object, o2 second object,
p non-verbal predicate, pp clause-level prepositional phrase, adv adverbial), measured directly
from alignment data, not asserted from grammar references. Every language aligns to the same Hebrew
source, so each aligned OT verse is a small parallel-order observation; aggregated over tens of thousands
of verses per language this yields real, per-language word-order fingerprints — validated against known
typology (Arabic preserves Hebrew's verb-first v>s order 91% of the time; English keeps it only 40%, i.e.
flips it to subject-first 60% of the time; Indonesian sits at 50%, consistent with its verb-initial narrative
register; Hindi keeps v>s 33% and v>o 39%, the signature of a verb-final language).
pair_order_kept: for each ADJACENT source-order constituent pair (a before b in the Hebrew), the share of
aligned verses where the target preserved that order. function_drift: mean normalized target-position minus
source-position per constituent label — which constituents a language systematically fronts or defers (the
key keeps its historical name; the labels are the phrase_role values above). Both are computed only from
cells with >=30 observations (see constituent_order.py). A constituent is a maximal run of consecutive
source tokens with the same clause (the verb they depend on) and phrase_role; its target position is the
mean of its aligned content tokens. Each file carries label_scheme: "macula_phrase_role".
Label vocabulary changed 2026-10-04. Earlier versions of these files used the ETCBC BHSA function
labels (Subj, Pred, Objc, ...). BHSA is licensed CC BY-NC-SA, so the data was re-derived from the
CC BY 4.0 MACULA treebank columns only and the BHSA-labelled files were withdrawn. The two vocabularies are
not interchangeable (7 labels instead of 14+), but the shared pairs agree closely (English verb-then-subject
kept 0.404 under BHSA, 0.403 under MACULA).
OT-only (the MACULA phrase layer is Hebrew-only). CC0-1.0 — derived alignment statistics, no source text redistributed.
- Downloads last month
- 69