Datasets:
The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
</tool_call>: int64
<tool_call>: int64
<|box_end|>: int64
<|box_start|>: int64
<|endoftext|>: int64
<|file_sep|>: int64
<|fim_middle|>: int64
<|fim_pad|>: int64
<|fim_prefix|>: int64
<|fim_suffix|>: int64
<|im_end|>: int64
<|im_start|>: int64
<|image_pad|>: int64
<|object_ref_end|>: int64
<|object_ref_start|>: int64
<|quad_end|>: int64
<|quad_start|>: int64
<|repo_name|>: int64
<|video_pad|>: int64
<|vision_end|>: int64
<|vision_pad|>: int64
<|vision_start|>: int64
scope: string
manifest_hash: string
checkpoint_commit_hash: string
version: int64
format: string
files: list<item: struct<path: string, bytes: int64, sha256: string>>
child 0, item: struct<path: string, bytes: int64, sha256: string>
child 0, path: string
child 1, bytes: int64
child 2, sha256: string
total_bytes: int64
to
{'format': Value('string'), 'version': Value('int64'), 'checkpoint_commit_hash': Value('string'), 'files': List({'path': Value('string'), 'bytes': Value('int64'), 'sha256': Value('string')}), 'total_bytes': Value('int64'), 'scope': Value('string'), 'manifest_hash': Value('string')}
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
</tool_call>: int64
<tool_call>: int64
<|box_end|>: int64
<|box_start|>: int64
<|endoftext|>: int64
<|file_sep|>: int64
<|fim_middle|>: int64
<|fim_pad|>: int64
<|fim_prefix|>: int64
<|fim_suffix|>: int64
<|im_end|>: int64
<|im_start|>: int64
<|image_pad|>: int64
<|object_ref_end|>: int64
<|object_ref_start|>: int64
<|quad_end|>: int64
<|quad_start|>: int64
<|repo_name|>: int64
<|video_pad|>: int64
<|vision_end|>: int64
<|vision_pad|>: int64
<|vision_start|>: int64
scope: string
manifest_hash: string
checkpoint_commit_hash: string
version: int64
format: string
files: list<item: struct<path: string, bytes: int64, sha256: string>>
child 0, item: struct<path: string, bytes: int64, sha256: string>
child 0, path: string
child 1, bytes: int64
child 2, sha256: string
total_bytes: int64
to
{'format': Value('string'), 'version': Value('int64'), 'checkpoint_commit_hash': Value('string'), 'files': List({'path': Value('string'), 'bytes': Value('int64'), 'sha256': Value('string')}), 'total_bytes': Value('int64'), 'scope': Value('string'), 'manifest_hash': 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.
SearchForge research artifacts
Versioned synthetic-world data, typed task programs, evidence audits, protected splits, and recovery/training states for the SearchForge symbolic task-model project. These artifacts support source-to-program-to-grounding reproduction.
The first snapshot is recovery and implementation evidence, not a successful solver-learning result. Historical findings remain: E3 PASS; E5a PASS; confirmatory E5 FAIL under its preregistered criterion; E6 INCONCLUSIVE. No symbolic SFT/static comparison is claimed until a corresponding run exists.
Snapshots and restoration
Each snapshots/<id>/ has a MANIFEST, gzip/tar parts and a COMPLETE marker
published together in one Hub commit. Pin the commit SHA, validate each part's
SHA-256, concatenate parts in manifest order, extract into a fresh directory,
and validate all restored file sizes and SHA-256 values. COMPLETE means complete
publication; a separate restoration receipt records actual state-loading checks.
Content includes a 28-type/84-relation synthetic world, 137-token IR, complete replay/grounding outcomes, protected splits, and bounded recovery-smoke states. Failures and costs are retained. Protected solver tests must not be used to train or tune generators. Original source and environment locks remain in the original GitHub repository; the manifest records its exact commit.
Upstream data: https://huggingface.co/datasets/hkuzxc/SearchGym-test-data, revision 5bf73cafe184610b40933ab2f42fb21c7926bd81.
Licensing
Third-party materials retain their upstream terms and provenance; this repository does not grant a blanket license for unrelated source repositories.
Built with Qwen. R4 derives from Qwen/Qwen2.5-3B revision
3aab1f1954e9cc14eb9509a215f9e5ca08227a9b. Its Qwen RESEARCH LICENSE AGREEMENT
and required notice are included in licenses/. Those materials are for
non-commercial research/evaluation under that agreement. The checkpoint was
reserialized by AReaL after one recovery-only AdamW step, creating optimizer
state. Model tensor hashes did not change because the group had zero utility.
It is not an old M4 or a formal training anchor.
Only explicitly selected research files are included. Account credentials, other user projects and whole private source archives are excluded. Public Hub storage is best-effort, not a guarantee of unlimited capacity or availability.
- Downloads last month
- 1,329