Datasets:
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
ts: double
variant: string
qid: string
pid: string
status: int64
latency_s: double
qhash: string
model: string
answer: struct<relevance: struct<type: string, score: double, legend: struct<0: string, 1: string, 2: string (... 101 chars omitted)
child 0, relevance: struct<type: string, score: double, legend: struct<0: string, 1: string, 2: string, 3: string>, prob (... 82 chars omitted)
child 0, type: string
child 1, score: double
child 2, legend: struct<0: string, 1: string, 2: string, 3: string>
child 0, 0: string
child 1, 1: string
child 2, 2: string
child 3, 3: string
child 3, probabilities: struct<0: double, 1: double, 2: double, 3: double>
child 0, 0: double
child 1, 1: double
child 2, 2: double
child 3, 3: double
child 4, confidence: double
usage: struct<input_tokens: int64, output_tokens: int64, cost: double>
child 0, input_tokens: int64
child 1, output_tokens: int64
child 2, cost: double
error: null
query: string
pool: list<item: struct<pid: string, text: string, relevant: bool>>
child 0, item: struct<pid: string, text: string, relevant: bool>
child 0, pid: string
child 1, text: string
child 2, relevant: bool
to
{'qid': Value('string'), 'query': Value('string'), 'pool': List({'pid': Value('string'), 'text': Value('string'), 'relevant': Value('bool')})}
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 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 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
ts: double
variant: string
qid: string
pid: string
status: int64
latency_s: double
qhash: string
model: string
answer: struct<relevance: struct<type: string, score: double, legend: struct<0: string, 1: string, 2: string (... 101 chars omitted)
child 0, relevance: struct<type: string, score: double, legend: struct<0: string, 1: string, 2: string, 3: string>, prob (... 82 chars omitted)
child 0, type: string
child 1, score: double
child 2, legend: struct<0: string, 1: string, 2: string, 3: string>
child 0, 0: string
child 1, 1: string
child 2, 2: string
child 3, 3: string
child 3, probabilities: struct<0: double, 1: double, 2: double, 3: double>
child 0, 0: double
child 1, 1: double
child 2, 2: double
child 3, 3: double
child 4, confidence: double
usage: struct<input_tokens: int64, output_tokens: int64, cost: double>
child 0, input_tokens: int64
child 1, output_tokens: int64
child 2, cost: double
error: null
query: string
pool: list<item: struct<pid: string, text: string, relevant: bool>>
child 0, item: struct<pid: string, text: string, relevant: bool>
child 0, pid: string
child 1, text: string
child 2, relevant: bool
to
{'qid': Value('string'), 'query': Value('string'), 'pool': List({'pid': Value('string'), 'text': Value('string'), 'relevant': Value('bool')})}
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.
Jev RAG Chinese Benchmark (S1)
First reproducible Chinese-language reranking measurements for Jev (TypeSafe System One), over hard negatives from C-MTEB/T2Reranking (T2Ranking lineage).
Headline (n=100 queries, pinned typesafe/jev-1.13-20260917, single run)
| Ranker | nDCG@5 (95% CI) | nDCG@10 | R@1 | R@5 |
|---|---|---|---|---|
| Jev Score (4-level rubric, zh) | 0.546 [0.481, 0.611] | 0.602 | 0.260 | 0.830 |
| Jev Noul (binary, zh) | 0.531 | 0.588 | 0.240 | 0.830 |
| BM25 (jieba) | 0.413 [0.353, 0.477] | 0.513 | 0.110 | 0.690 |
Paired delta (Jev Score − BM25): nDCG@5 +0.133 [95% CI +0.051, +0.232] — bootstrap over queries, CI excludes zero. Total API cost: $0.07.
Contents
s1.jsonl/dry.jsonl— sampled pools (100 / 20 queries; 1 positive + 9 hard negatives each, seed 42)logs/— every API call: timestamp, variant, latency, frozen-question hash, returned model id, answer, usage/costresults/— full report with bootstrap CIs and the Noul threshold sweepPROTOCOL.md— the frozen protocol
Code, scripts, and the arXiv technical-note draft: github.com/vicfei/jev-rag-zh-benchmark · companion pattern library: awesome-jev-prompts.
Independent community research; not affiliated with TypeSafe AI. Single run per variant; limitations are documented in the protocol and the report.
- Downloads last month
- 49