The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: TypeError
Message: Couldn't cast array of type
struct<repository: string, revision: string, field: string, subset: string, license_note: string, files_read: list<item: string>, start_row: int64, phase: string, selection: string, identity_sha256: string, implementation_sha256: string, end_row_exclusive: int64, file_exhausted: bool, filter_counts: struct<accepted: int64, exact: int64, near: int64, benchmark: int64, short_or_long: int64, discarded_tiny_tail_tokens: int64>>
to
{'repository': Value('string'), 'revision': Value('string'), 'field': Value('string'), 'subset': Value('string'), 'license_note': Value('string'), 'files_read': List(Value('string')), 'start_row': Value('int64'), 'end_row_exclusive': Value('int64'), 'file_exhausted': Value('bool'), 'phase': Value('string'), 'selection': Value('string'), 'identity_sha256': Value('string'), 'execution_migration': Value('string'), 'filter_counts': {'accepted': Value('int64'), 'exact': Value('int64'), 'near': Value('int64'), 'benchmark': Value('int64'), 'short_or_long': Value('int64')}}
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 2315, 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 2161, 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<repository: string, revision: string, field: string, subset: string, license_note: string, files_read: list<item: string>, start_row: int64, phase: string, selection: string, identity_sha256: string, implementation_sha256: string, end_row_exclusive: int64, file_exhausted: bool, filter_counts: struct<accepted: int64, exact: int64, near: int64, benchmark: int64, short_or_long: int64, discarded_tiny_tail_tokens: int64>>
to
{'repository': Value('string'), 'revision': Value('string'), 'field': Value('string'), 'subset': Value('string'), 'license_note': Value('string'), 'files_read': List(Value('string')), 'start_row': Value('int64'), 'end_row_exclusive': Value('int64'), 'file_exhausted': Value('bool'), 'phase': Value('string'), 'selection': Value('string'), 'identity_sha256': Value('string'), 'execution_migration': Value('string'), 'filter_counts': {'accepted': Value('int64'), 'exact': Value('int64'), 'near': Value('int64'), 'benchmark': Value('int64'), 'short_or_long': 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.
Private Ivme mixture for sub-150M experiments
Status: PREPARING. This dataset is private.
Preparation releases balanced 1B and 4B training prefixes before the full 20B target. Only published screening manifests or the completed train/manifest.json are ready to train. Prepared tokens are distinct filtered document material, not repeated training exposures. Exact and conservative near deduplication are applied; this is not a guarantee of unique passages.
Ready training manifests
- 1000000000: 1,000,000,000 tokens;
screening/1000000000/manifest.jsonat revision54cd2447637decf7c3ed9d410ee01d96a4b8b0fa.
Evaluation separation
Never train on final-holdout/ or final-holdout-batches/. The new 1M final holdout uses reserved files and is indexed for duplicate exclusion before training. Use the earlier development holdout for recipe selection. The untrained V2 100M pilot is pipeline validation only. Cached BLiMP, ARC, WikiText-2, HellaSwag, PIQA and WinoGrande text is used only as a contamination exclusion index. Exact overlap screening does not certify absence of semantic contamination.
Tokenizer and mixture
Frozen vocabulary: 16,384. Tokenizer SHA-256: a847369bb0361ae4562e87d53357f47ddfb6b52fa9e1b8d00fc68d6207022e81.
Initial shares: FineWeb-Edu46.67%, DCLM27.78%, FineWiki8.89%, FineMath7.78%, Cosmopedia5.56%, Stories3.32%.
If distinct Stories is exhausted, its measured deficit is assigned to FineWeb-Edu and actual shares are recorded.
The tokenizer was fitted to a small starter sample and still needs representative compression review.
Sources and licensing
Upstream notices and licensing obligations remain applicable; mixed-source metadata is not a single blanket license.
- fineweb_edu: odc-by.
- dclm: cc-by-4.0.
- finewiki: cc-by-sa-4.0 / gfdl; review upstream terms.
- finemath: odc-by.
- cosmopedia: odc-by.
- stories: mit.
Pinned files, filtering rules and benchmark revisions are under metadata/. Frozen code is under implementation/. Verified upload receipts precede local payload eviction. Immutable screening manifests preserve their exact shards. Full completion is recorded under completion/progress.json; intermediate status snapshots are under progress-snapshots/. No SOTA or benchmark-quality claim is made for this preparation.
- Downloads last month
- 2