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: CastError
Message: Couldn't cast
family: string
input: struct<arrays: struct<document_embeddings: struct<dtype: string, shape: list<item: int64>>, document (... 1039 chars omitted)
child 0, arrays: struct<document_embeddings: struct<dtype: string, shape: list<item: int64>>, document_offsets: struc (... 299 chars omitted)
child 0, document_embeddings: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 1, document_offsets: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 2, query_embeddings: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 3, query_offsets: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 4, sample_groups: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 5, sample_ids: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 1, export_manifest: struct<path: string, probe_mani
...
d 2, sha256: string
runtime: struct<cuda: string, gpu_name: string, pylate: string, sentence_transformers: string, torch: string>
child 0, cuda: string
child 1, gpu_name: string
child 2, pylate: string
child 3, sentence_transformers: string
child 4, torch: string
probe: struct<dataset_fingerprint: string, frozen_spec: struct<expected_manifest_sha256: string, path: stri (... 121 chars omitted)
child 0, dataset_fingerprint: string
child 1, frozen_spec: struct<expected_manifest_sha256: string, path: string, sha256: string>
child 0, expected_manifest_sha256: string
child 1, path: string
child 2, sha256: string
child 2, manifest_sha256: string
child 3, path: string
child 4, selected_sample_ids_sha256: string
child 5, selection_sha256: string
checkpoint_run_config: struct<path: string, sha256: string>
child 0, path: string
child 1, sha256: string
encoding: struct<batch_size: int64, compressed: bool, dense_document_prompt: null, dense_query_prompt: null, d (... 231 chars omitted)
child 0, batch_size: int64
child 1, compressed: bool
child 2, dense_document_prompt: null
child 3, dense_query_prompt: null
child 4, device: string
child 5, flash_attention: bool
child 6, late_document_skiplist: bool
child 7, late_query_expansion: bool
child 8, late_storage: string
child 9, max_length: int64
child 10, model_dtype: string
child 11, normalized: bool
child 12, positive_candidate_index: int64
child 13, storage_dtype: string
to
{'checkpoint': Value('string'), 'checkpoint_inputs': List({'bytes': Value('int64'), 'path': Value('string'), 'sha256': Value('string')}), 'checkpoint_run_config': {'path': Value('string'), 'sha256': Value('string')}, 'encoding': {'batch_size': Value('int64'), 'compressed': Value('bool'), 'dense_document_prompt': Value('null'), 'dense_query_prompt': Value('null'), 'device': Value('string'), 'flash_attention': Value('bool'), 'late_document_skiplist': Value('bool'), 'late_query_expansion': Value('bool'), 'late_storage': Value('string'), 'max_length': Value('int64'), 'model_dtype': Value('string'), 'normalized': Value('bool'), 'positive_candidate_index': Value('int64'), 'storage_dtype': Value('string')}, 'family': Value('string'), 'output': {'arrays': {'document_embeddings': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'document_offsets': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'query_embeddings': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'query_offsets': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'sample_groups': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'sample_ids': {'dtype': Value('string'), 'shape': List(Value('int64'))}}, 'bytes': Value('int64'), 'path': Value('string'), 'sha256': Value('string')}, 'probe': {'dataset_fingerprint': Value('string'), 'frozen_spec': {'expected_manifest_sha256': Value('string'), 'path': Value('string'), 'sha256': Value('string')}, 'manifest_sha256': Value('string'), 'path': Value('string'), 'selected_sample_ids_sha256': Value('string'), 'selection_sha256': Value('string')}, 'runtime': {'cuda': Value('string'), 'gpu_name': Value('string'), 'pylate': Value('string'), 'sentence_transformers': Value('string'), 'torch': Value('string')}, 'schema_version': Value('int64')}
because column names don't match
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 2840, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, 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 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, 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 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
family: string
input: struct<arrays: struct<document_embeddings: struct<dtype: string, shape: list<item: int64>>, document (... 1039 chars omitted)
child 0, arrays: struct<document_embeddings: struct<dtype: string, shape: list<item: int64>>, document_offsets: struc (... 299 chars omitted)
child 0, document_embeddings: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 1, document_offsets: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 2, query_embeddings: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 3, query_offsets: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 4, sample_groups: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 5, sample_ids: struct<dtype: string, shape: list<item: int64>>
child 0, dtype: string
child 1, shape: list<item: int64>
child 0, item: int64
child 1, export_manifest: struct<path: string, probe_mani
...
d 2, sha256: string
runtime: struct<cuda: string, gpu_name: string, pylate: string, sentence_transformers: string, torch: string>
child 0, cuda: string
child 1, gpu_name: string
child 2, pylate: string
child 3, sentence_transformers: string
child 4, torch: string
probe: struct<dataset_fingerprint: string, frozen_spec: struct<expected_manifest_sha256: string, path: stri (... 121 chars omitted)
child 0, dataset_fingerprint: string
child 1, frozen_spec: struct<expected_manifest_sha256: string, path: string, sha256: string>
child 0, expected_manifest_sha256: string
child 1, path: string
child 2, sha256: string
child 2, manifest_sha256: string
child 3, path: string
child 4, selected_sample_ids_sha256: string
child 5, selection_sha256: string
checkpoint_run_config: struct<path: string, sha256: string>
child 0, path: string
child 1, sha256: string
encoding: struct<batch_size: int64, compressed: bool, dense_document_prompt: null, dense_query_prompt: null, d (... 231 chars omitted)
child 0, batch_size: int64
child 1, compressed: bool
child 2, dense_document_prompt: null
child 3, dense_query_prompt: null
child 4, device: string
child 5, flash_attention: bool
child 6, late_document_skiplist: bool
child 7, late_query_expansion: bool
child 8, late_storage: string
child 9, max_length: int64
child 10, model_dtype: string
child 11, normalized: bool
child 12, positive_candidate_index: int64
child 13, storage_dtype: string
to
{'checkpoint': Value('string'), 'checkpoint_inputs': List({'bytes': Value('int64'), 'path': Value('string'), 'sha256': Value('string')}), 'checkpoint_run_config': {'path': Value('string'), 'sha256': Value('string')}, 'encoding': {'batch_size': Value('int64'), 'compressed': Value('bool'), 'dense_document_prompt': Value('null'), 'dense_query_prompt': Value('null'), 'device': Value('string'), 'flash_attention': Value('bool'), 'late_document_skiplist': Value('bool'), 'late_query_expansion': Value('bool'), 'late_storage': Value('string'), 'max_length': Value('int64'), 'model_dtype': Value('string'), 'normalized': Value('bool'), 'positive_candidate_index': Value('int64'), 'storage_dtype': Value('string')}, 'family': Value('string'), 'output': {'arrays': {'document_embeddings': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'document_offsets': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'query_embeddings': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'query_offsets': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'sample_groups': {'dtype': Value('string'), 'shape': List(Value('int64'))}, 'sample_ids': {'dtype': Value('string'), 'shape': List(Value('int64'))}}, 'bytes': Value('int64'), 'path': Value('string'), 'sha256': Value('string')}, 'probe': {'dataset_fingerprint': Value('string'), 'frozen_spec': {'expected_manifest_sha256': Value('string'), 'path': Value('string'), 'sha256': Value('string')}, 'manifest_sha256': Value('string'), 'path': Value('string'), 'selected_sample_ids_sha256': Value('string'), 'selection_sha256': Value('string')}, 'runtime': {'cuda': Value('string'), 'gpu_name': Value('string'), 'pylate': Value('string'), 'sentence_transformers': Value('string'), 'torch': Value('string')}, 'schema_version': Value('int64')}
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.
Embedding optimizer study analysis artifacts
This repository preserves analysis artifacts for the current DenseOn comparison of AdamW, Muon and NorMuon. The source repository and paper contain the completed experiments, protocols, exact analysis and restoration tools. Model and optimizer states are in the separate checkpoint repository.
Current scientific artifacts
Use the immutable revisions and manifests in these guides, not a broad download
of the mixed-history repository. The current scientific namespace is
corrected-dense-correctness-v3/.
- Primary retrieval trajectories: all sixty primary checkpoints across fourteen BEIR tasks.
- Weight analyses.
- Functional vectors and calibration: 624 files, anonymously downloaded and checksum-verified.
- Continuation probes and checkpoint links: 638 files, including all sixty five-stage continuation probes.
- Continuation retrieval outcomes: 593 files covering all 168 final full-corpus task results and associated records.
Eight-candidate functional probes are diagnostics, not full-corpus compressed retrieval. Twelve primary learning-rate configurations are not twelve independent training seeds; the continuation-order seeds do not replicate source training. See the paper for supported conclusions and their scope.
The older corrected-dense-no-packing-v1/ namespace is not the current v3 study.
Retention of any historical file or configuration does not make its results valid
or current. Runtime artifacts and separately identified shared inputs are preserved.
Owner-approved withdrawal of obsolete results
This commit removes 5,839 explicitly reviewed old files from the current tree, following the owner's confirmation. The targets comprise invalidated Dense retrieval, representation and weight-space outputs, affected follow-up analyses, mixed result summaries and associated historical reports and execution logs. Their former values must not be cited as results of the current study.
The exact reviewed deletion manifest
has SHA-256 1193dbc2a1a483cba79b3532b4ef47f66eb9198fefc7492ea4521a9d00f1d026.
The commit changes only those paths and this README. Git history, LFS objects,
original immutable restoration references and local engineering records are not
deleted. Old revisions can still contain withdrawn files; availability is not
scientific validity. Shared objects remain intact.
Shared project/data/, historical configurations, independently encoded untrained
Dense BEIR baselines and separately identified LateOn artifacts are retained.
LateOn remains outside the current paper; retaining it is not a correctness claim.
This repository is no longer a complete mirror of an old local results/ folder.
Do not re-upload withdrawn files from broad historical snapshots.
- Downloads last month
- 788