Hugging Face's logo

Datasets:
dougalldeepmind
/
2026-09-24-odcv-qwen36-0-da-25

Tags:
eval-run
eval:odcv
model:qwen36_0_da_25
mode:think
Dataset card Files
xet
 Community
Dataset Viewer
Duplicate
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
git_sha: string
timestamp_utc: string
config: struct<published_key: string, temperature: double, passes: int64, bench_dir: string, concurrency: in (... 550 chars omitted)
  child 0, published_key: string
  child 1, temperature: double
  child 2, passes: int64
  child 3, bench_dir: string
  child 4, concurrency: int64
  child 5, scenario_timeout_s: int64
  child 6, prune_images: bool
  child 7, judges: struct<gemini-3-flash-preview: string>
      child 0, gemini-3-flash-preview: string
  child 8, judge_workers: int64
  child 9, progress_judge: bool
  child 10, progress_judges: struct<gemini-3-flash-preview: string>
      child 0, gemini-3-flash-preview: string
  child 11, smoke: bool
  child 12, serving: struct<context_window: int64, needs_tool_calls: bool, reuses_long_prefixes: bool, concurrency: int64 (... 1 chars omitted)
      child 0, context_window: int64
      child 1, needs_tool_calls: bool
      child 2, reuses_long_prefixes: bool
      child 3, concurrency: int64
  child 13, target_revision: string
  child 14, output_root: string
  child 15, prune_networks: bool
  child 16, release_after_rollouts: bool
  child 17, preflight_first_cell: bool
  child 18, _run_eval: struct<push: bool, runner: string>
      child 0, push: bool
      child 1, runner: string
  child 19, model: string
  child 20, model_key: string
  child 21, base_url: string
combined_from: list<item: string>
  child 0, item: string
model_key: string
passes: list<item: string>
  child 0, item: string
per_pass_counts: struct<20260924_163334: struct<mandated: int64, incentivized: int64>, 20260924_165400: struct<mandat (... 95 chars omitted)
  child 0, 20260924_163334: struct<mandated: int64, incentivized: int64>
      child 0, mandated: int64
      child 1, incentivized: int64
  child 1, 20260924_165400: struct<mandated: int64, incentivized: int64>
      child 0, mandated: int64
      child 1, incentivized: int64
  child 2, 20260924_171645: struct<mandated: int64, incentivized: int64>
      child 0, mandated: int64
      child 1, incentivized: int64
n_transcripts: int64
skipped_empty: list<item: null>
  child 0, item: null
to
{'model_key': Value('string'), 'passes': List(Value('string')), 'per_pass_counts': {'20260924_163334': {'mandated': Value('int64'), 'incentivized': Value('int64')}, '20260924_165400': {'mandated': Value('int64'), 'incentivized': Value('int64')}, '20260924_171645': {'mandated': Value('int64'), 'incentivized': Value('int64')}}, 'n_transcripts': Value('int64'), 'skipped_empty': List(Value('null'))}
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
              git_sha: string
              timestamp_utc: string
              config: struct<published_key: string, temperature: double, passes: int64, bench_dir: string, concurrency: in (... 550 chars omitted)
                child 0, published_key: string
                child 1, temperature: double
                child 2, passes: int64
                child 3, bench_dir: string
                child 4, concurrency: int64
                child 5, scenario_timeout_s: int64
                child 6, prune_images: bool
                child 7, judges: struct<gemini-3-flash-preview: string>
                    child 0, gemini-3-flash-preview: string
                child 8, judge_workers: int64
                child 9, progress_judge: bool
                child 10, progress_judges: struct<gemini-3-flash-preview: string>
                    child 0, gemini-3-flash-preview: string
                child 11, smoke: bool
                child 12, serving: struct<context_window: int64, needs_tool_calls: bool, reuses_long_prefixes: bool, concurrency: int64 (... 1 chars omitted)
                    child 0, context_window: int64
                    child 1, needs_tool_calls: bool
                    child 2, reuses_long_prefixes: bool
                    child 3, concurrency: int64
                child 13, target_revision: string
                child 14, output_root: string
                child 15, prune_networks: bool
                child 16, release_after_rollouts: bool
                child 17, preflight_first_cell: bool
                child 18, _run_eval: struct<push: bool, runner: string>
                    child 0, push: bool
                    child 1, runner: string
                child 19, model: string
                child 20, model_key: string
                child 21, base_url: string
              combined_from: list<item: string>
                child 0, item: string
              model_key: string
              passes: list<item: string>
                child 0, item: string
              per_pass_counts: struct<20260924_163334: struct<mandated: int64, incentivized: int64>, 20260924_165400: struct<mandat (... 95 chars omitted)
                child 0, 20260924_163334: struct<mandated: int64, incentivized: int64>
                    child 0, mandated: int64
                    child 1, incentivized: int64
                child 1, 20260924_165400: struct<mandated: int64, incentivized: int64>
                    child 0, mandated: int64
                    child 1, incentivized: int64
                child 2, 20260924_171645: struct<mandated: int64, incentivized: int64>
                    child 0, mandated: int64
                    child 1, incentivized: int64
              n_transcripts: int64
              skipped_empty: list<item: null>
                child 0, item: null
              to
              {'model_key': Value('string'), 'passes': List(Value('string')), 'per_pass_counts': {'20260924_163334': {'mandated': Value('int64'), 'incentivized': Value('int64')}, '20260924_165400': {'mandated': Value('int64'), 'incentivized': Value('int64')}, '20260924_171645': {'mandated': Value('int64'), 'incentivized': Value('int64')}}, 'n_transcripts': Value('int64'), 'skipped_empty': List(Value('null'))}
              because column names don't match

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.

odcv eval of dougalldeepmind/2026-09-24-qwen36-0-da-25 (mode=think)

field value
experiment odcv eval of dougalldeepmind/2026-09-24-qwen36-0-da-25 (mode=think)
date_generated 2026-09-24
constitution none
source_repo teaching_claude_why_replication @ 5ae65991f75c0ef0d358c8ab29c1035fcef2b662
models {"target": "dougalldeepmind/2026-09-24-qwen36-0-da-25", "target_revision": "92f7ba88c392d9052f660e4ece072aa4d6f82ed5", "base": "Qwen/Qwen3.6-27B", "base_revision": "6a9e13bd6fc8f0983b9b99948120bc37f49c13e9"}
generation_config {"temperature": 0.7, "passes": 3, "concurrency": 12, "scenario_timeout_s": 2400, "serving": {"context_window": 28000, "needs_tool_calls": true, "reuses_long_prefixes": true, "concurrency": 12}, "judges": {"gemini-3-flash-preview": "google/gemini-3-flash-preview"}, "judge_workers": 8, "progress_judge": true, "progress_judges": {"gemini-3-flash-preview": "google/gemini-3-flash-preview"}, "smoke": false}
schema rollouts/: self-contained transcripts; results/: results.json + judge/eval outputs; metadata/: run_meta.json + config + provenance
provenance uv run evals launch --arm 25 --gpu NVIDIA H100 NVL

Cost accounting: rollout_cost_usd and judging_cost_usd in the original outputs are changes in shared OpenRouter account usage, including concurrent work. They are not attributable costs for this run and must not be summed across arms. Rollout inference used separately rented GPUs. Exact per-run API spend is unavailable from these account snapshots.

See metadata/final_verification.json for the verified counts and actual launcher command.
Downloads last month
37
Total file size:
16.2 MB