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
codec: string
payload: string
schema: string
policies: struct<lunatic_reimu_a_stage6.json: struct<bytes: int64, coarse_action_states: int64, decisions: int (... 231 chars omitted)
child 0, lunatic_reimu_a_stage6.json: struct<bytes: int64, coarse_action_states: int64, decisions: int64, exploratory_decisions: int64, fi (... 194 chars omitted)
child 0, bytes: int64
child 1, coarse_action_states: int64
child 2, decisions: int64
child 3, exploratory_decisions: int64
child 4, fine_action_states: int64
child 5, middle_action_states: int64
child 6, observed_trials: int64
child 7, outer_schema: string
child 8, path: string
child 9, reward_version: string
child 10, selection: string
child 11, sha256: string
child 12, state_schema: string
snapshot_id: string
to
{'policies': {'lunatic_reimu_a_stage6.json': {'bytes': Value('int64'), 'coarse_action_states': Value('int64'), 'decisions': Value('int64'), 'exploratory_decisions': Value('int64'), 'fine_action_states': Value('int64'), 'middle_action_states': Value('int64'), 'observed_trials': Value('int64'), 'outer_schema': Value('string'), 'path': Value('string'), 'reward_version': Value('string'), 'selection': Value('string'), 'sha256': Value('string'), 'state_schema': Value('string')}}, 'schema': Value('string'), 'snapshot_id': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, 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 129, 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 489, 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 2818, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2355, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2380, 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 2369, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
codec: string
payload: string
schema: string
policies: struct<lunatic_reimu_a_stage6.json: struct<bytes: int64, coarse_action_states: int64, decisions: int (... 231 chars omitted)
child 0, lunatic_reimu_a_stage6.json: struct<bytes: int64, coarse_action_states: int64, decisions: int64, exploratory_decisions: int64, fi (... 194 chars omitted)
child 0, bytes: int64
child 1, coarse_action_states: int64
child 2, decisions: int64
child 3, exploratory_decisions: int64
child 4, fine_action_states: int64
child 5, middle_action_states: int64
child 6, observed_trials: int64
child 7, outer_schema: string
child 8, path: string
child 9, reward_version: string
child 10, selection: string
child 11, sha256: string
child 12, state_schema: string
snapshot_id: string
to
{'policies': {'lunatic_reimu_a_stage6.json': {'bytes': Value('int64'), 'coarse_action_states': Value('int64'), 'decisions': Value('int64'), 'exploratory_decisions': Value('int64'), 'fine_action_states': Value('int64'), 'middle_action_states': Value('int64'), 'observed_trials': Value('int64'), 'outer_schema': Value('string'), 'path': Value('string'), 'reward_version': Value('string'), 'selection': Value('string'), 'sha256': Value('string'), 'state_schema': Value('string')}}, 'schema': Value('string'), 'snapshot_id': 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.
TH06 source-grounded RL corpus
This private, append-only dataset mirrors physical trajectories produced by touhou-solver-th06-rl. Difficulty, character, shot type, Stage, and automatically derived source phase remain explicit learning scopes.
runs/<run_id>/ preserves recorder-native gzip shards plus their immutable
run.json, manifest.json, and optional infrastructure audit. The dataset
index labels complete-stage/training eligibility; diagnostic and legacy runs
are retained but must not silently enter that stratum. checkpoints/ contains
content-addressed snapshots of the last Stage-committed online sufficient
statistics. A checkpoint is deployment state, not a replacement for raw data
when training a new offline model or reward definition.
The corpus contains decoded gameplay state and outcomes, not the game executable, DAT assets, screenshots, credentials, or process-memory dumps. Bomb remains forbidden and every deployed action remains subject to the native safety gate. Observational trajectories are not counterfactual proof; physical complete-Stage play is the final evaluation.
- Downloads last month
- 603