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: ArrowInvalid
Message: Could not open Parquet input source '<Buffer>': Parquet magic bytes not found in footer. Either the file is corrupted or this is not a parquet file.
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/parquet/parquet.py", line 205, in _generate_tables
if parquet_fragment.row_groups:
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "pyarrow/_dataset_parquet.pyx", line 389, in pyarrow._dataset_parquet.ParquetFileFragment.row_groups.__get__
File "pyarrow/_dataset_parquet.pyx", line 396, in pyarrow._dataset_parquet.ParquetFileFragment.metadata.__get__
self.ensure_complete_metadata()
File "pyarrow/_dataset_parquet.pyx", line 385, in pyarrow._dataset_parquet.ParquetFileFragment.ensure_complete_metadata
check_status(self.parquet_file_fragment.EnsureCompleteMetadata())
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
pyarrow.lib.ArrowInvalid: Could not open Parquet input source '<Buffer>': Parquet magic bytes not found in footer. Either the file is corrupted or this is not a parquet file.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.
LIMA Check — Retrofitted Explicit Reasoning
This dataset contains 139 examples from essobi/lima_check that have been retrofitted with explicit reasoning traces.
Description
The original lima_check dataset contains StackExchange and other Q&A examples categorized by reasoning style:
- No Reasoning (876 examples)
- Implicit Reasoning (153 examples) — reasoning embedded in the answer without explicit structure
- Explicit Reasoning (1 example)
This subset contains the 139 successfully retrofitted examples where implicit reasoning traces were extracted using a teacher model (coder-model / Qwen3.6-35B) and wrapped in <reasoning> tags.
Reasoning Categories
| Category | Count |
|---|---|
| Retrofitted Explicit Reasoning | 139 |
Sources
| Source | Count |
|---|---|
| stackexchange | 110 |
| authors | 13 |
| nlp | 9 |
| multi_turn | 9 |
| wikihow | 4 |
Format
Each example contains:
- conversations: JSON string with
user/assistantmessage pairs. The assistant message contains<reasoning>...</reasoning>followed by the original answer. - source: Original data source
- reasoning_category: Always
Retrofitted Explicit Reasoning - reasoning_confidence: Confidence level of the original annotation
- operations: List of reasoning operations applied
- flags: Boolean flags for reasoning operations
Usage
Load with datasets:
from datasets import load_dataset
ds = load_dataset("essobi/lima_check_retrofitted")
Script
The retrofitting was done with retrofit_reasoning.py in the mcp-dataset-engineering repo.
- Downloads last month
- 35