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
bddl: struct<bytes: int64, path: string, sha256: string>
child 0, bytes: int64
child 1, path: string
child 2, sha256: string
init_state_ids: list<item: int64>
child 0, item: int64
init_states: struct<bytes: int64, dtype: string, path: string, serialization: string, sha256: string, shape: list (... 14 chars omitted)
child 0, bytes: int64
child 1, dtype: string
child 2, path: string
child 3, serialization: string
child 4, sha256: string
child 5, shape: list<item: int64>
child 0, item: int64
instruction: string
rollout_count: int64
schema_version: string
suite: string
suite_index: int64
task_id: int64
task_name: string
protocol: struct<control_hz: int64, early_success_terminates_rollout: bool, evaluation_mode: string, fastwam_r (... 407 chars omitted)
child 0, control_hz: int64
child 1, early_success_terminates_rollout: bool
child 2, evaluation_mode: string
child 3, fastwam_render_resolution: int64
child 4, fastwam_replan_steps: int64
child 5, reset_order: list<item: string>
child 0, item: string
child 6, seed: int64
child 7, startup_wait: struct<counts_toward_policy_step_budget: bool, dummy_action_env: list<item: double>, steps: int64>
child 0, counts_toward_policy_step_budget: bool
child 1, dummy_action_env: list<item: double>
child 0, item: double
child 2, steps: int64
child 8, step_budget_by_suite: struct<libero_10: int64, libero_goal: int64, libero_object: int64, libero_spatial: int64>
child
...
rollouts: int64
child 1, tasks: int64
child 1, libero_goal: struct<rollouts: int64, tasks: int64>
child 0, rollouts: int64
child 1, tasks: int64
child 2, libero_object: struct<rollouts: int64, tasks: int64>
child 0, rollouts: int64
child 1, tasks: int64
child 3, libero_spatial: struct<rollouts: int64, tasks: int64>
child 0, rollouts: int64
child 1, tasks: int64
child 1, init_state_ids: list<item: int64>
child 0, item: int64
child 2, rollouts: int64
child 3, states_per_task: int64
child 4, suite_order: list<item: string>
child 0, item: string
child 5, task_ids: list<item: int64>
child 0, item: int64
child 6, tasks: int64
inventory: struct<bddl: struct<bytes: int64, files: int64>, init_states: struct<arrays: int64, bytes: int64, fi (... 148 chars omitted)
child 0, bddl: struct<bytes: int64, files: int64>
child 0, bytes: int64
child 1, files: int64
child 1, init_states: struct<arrays: int64, bytes: int64, files: int64, rows: int64>
child 0, arrays: int64
child 1, bytes: int64
child 2, files: int64
child 3, rows: int64
child 2, payload: struct<bytes: int64, files: int64>
child 0, bytes: int64
child 1, files: int64
child 3, tasks_jsonl: struct<bytes: int64, path: string, rows: int64, sha256: string>
child 0, bytes: int64
child 1, path: string
child 2, rows: int64
child 3, sha256: string
to
{'inventory': {'bddl': {'bytes': Value('int64'), 'files': Value('int64')}, 'init_states': {'arrays': Value('int64'), 'bytes': Value('int64'), 'files': Value('int64'), 'rows': Value('int64')}, 'payload': {'bytes': Value('int64'), 'files': Value('int64')}, 'tasks_jsonl': {'bytes': Value('int64'), 'path': Value('string'), 'rows': Value('int64'), 'sha256': Value('string')}}, 'population': {'by_suite': {'libero_10': {'rollouts': Value('int64'), 'tasks': Value('int64')}, 'libero_goal': {'rollouts': Value('int64'), 'tasks': Value('int64')}, 'libero_object': {'rollouts': Value('int64'), 'tasks': Value('int64')}, 'libero_spatial': {'rollouts': Value('int64'), 'tasks': Value('int64')}}, 'init_state_ids': List(Value('int64')), 'rollouts': Value('int64'), 'states_per_task': Value('int64'), 'suite_order': List(Value('string')), 'task_ids': List(Value('int64')), 'tasks': Value('int64')}, 'protocol': {'control_hz': Value('int64'), 'early_success_terminates_rollout': Value('bool'), 'evaluation_mode': Value('string'), 'fastwam_render_resolution': Value('int64'), 'fastwam_replan_steps': Value('int64'), 'reset_order': List(Value('string')), 'seed': Value('int64'), 'startup_wait': {'counts_toward_policy_step_budget': Value('bool'), 'dummy_action_env': List(Value('float64')), 'steps': Value('int64')}, 'step_budget_by_suite': {'libero_10': Value('int64'), 'libero_goal': Value('int64'), 'libero_object': Value('int64'), 'libero_spatial': Value('int64')}, 'task_environment_lifetime': Value('string'), 'task_order_index': Value('int64'), 'trial_order': Value('string')}, 'schema_version': Value('string'), 'source': {'libero_commit': Value('string'), 'relevant_paths_clean': Value('bool'), 'task_map': {'bytes': Value('int64'), 'path': Value('string'), 'sha256': Value('string')}}, 'status': Value('string'), 'tree_hashes': {'algorithm': Value('string'), 'bddl_tree_sha256': Value('string'), 'format': Value('string'), 'init_states_tree_sha256': Value('string'), 'payload_tree_sha256': 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
bddl: struct<bytes: int64, path: string, sha256: string>
child 0, bytes: int64
child 1, path: string
child 2, sha256: string
init_state_ids: list<item: int64>
child 0, item: int64
init_states: struct<bytes: int64, dtype: string, path: string, serialization: string, sha256: string, shape: list (... 14 chars omitted)
child 0, bytes: int64
child 1, dtype: string
child 2, path: string
child 3, serialization: string
child 4, sha256: string
child 5, shape: list<item: int64>
child 0, item: int64
instruction: string
rollout_count: int64
schema_version: string
suite: string
suite_index: int64
task_id: int64
task_name: string
protocol: struct<control_hz: int64, early_success_terminates_rollout: bool, evaluation_mode: string, fastwam_r (... 407 chars omitted)
child 0, control_hz: int64
child 1, early_success_terminates_rollout: bool
child 2, evaluation_mode: string
child 3, fastwam_render_resolution: int64
child 4, fastwam_replan_steps: int64
child 5, reset_order: list<item: string>
child 0, item: string
child 6, seed: int64
child 7, startup_wait: struct<counts_toward_policy_step_budget: bool, dummy_action_env: list<item: double>, steps: int64>
child 0, counts_toward_policy_step_budget: bool
child 1, dummy_action_env: list<item: double>
child 0, item: double
child 2, steps: int64
child 8, step_budget_by_suite: struct<libero_10: int64, libero_goal: int64, libero_object: int64, libero_spatial: int64>
child
...
rollouts: int64
child 1, tasks: int64
child 1, libero_goal: struct<rollouts: int64, tasks: int64>
child 0, rollouts: int64
child 1, tasks: int64
child 2, libero_object: struct<rollouts: int64, tasks: int64>
child 0, rollouts: int64
child 1, tasks: int64
child 3, libero_spatial: struct<rollouts: int64, tasks: int64>
child 0, rollouts: int64
child 1, tasks: int64
child 1, init_state_ids: list<item: int64>
child 0, item: int64
child 2, rollouts: int64
child 3, states_per_task: int64
child 4, suite_order: list<item: string>
child 0, item: string
child 5, task_ids: list<item: int64>
child 0, item: int64
child 6, tasks: int64
inventory: struct<bddl: struct<bytes: int64, files: int64>, init_states: struct<arrays: int64, bytes: int64, fi (... 148 chars omitted)
child 0, bddl: struct<bytes: int64, files: int64>
child 0, bytes: int64
child 1, files: int64
child 1, init_states: struct<arrays: int64, bytes: int64, files: int64, rows: int64>
child 0, arrays: int64
child 1, bytes: int64
child 2, files: int64
child 3, rows: int64
child 2, payload: struct<bytes: int64, files: int64>
child 0, bytes: int64
child 1, files: int64
child 3, tasks_jsonl: struct<bytes: int64, path: string, rows: int64, sha256: string>
child 0, bytes: int64
child 1, path: string
child 2, rows: int64
child 3, sha256: string
to
{'inventory': {'bddl': {'bytes': Value('int64'), 'files': Value('int64')}, 'init_states': {'arrays': Value('int64'), 'bytes': Value('int64'), 'files': Value('int64'), 'rows': Value('int64')}, 'payload': {'bytes': Value('int64'), 'files': Value('int64')}, 'tasks_jsonl': {'bytes': Value('int64'), 'path': Value('string'), 'rows': Value('int64'), 'sha256': Value('string')}}, 'population': {'by_suite': {'libero_10': {'rollouts': Value('int64'), 'tasks': Value('int64')}, 'libero_goal': {'rollouts': Value('int64'), 'tasks': Value('int64')}, 'libero_object': {'rollouts': Value('int64'), 'tasks': Value('int64')}, 'libero_spatial': {'rollouts': Value('int64'), 'tasks': Value('int64')}}, 'init_state_ids': List(Value('int64')), 'rollouts': Value('int64'), 'states_per_task': Value('int64'), 'suite_order': List(Value('string')), 'task_ids': List(Value('int64')), 'tasks': Value('int64')}, 'protocol': {'control_hz': Value('int64'), 'early_success_terminates_rollout': Value('bool'), 'evaluation_mode': Value('string'), 'fastwam_render_resolution': Value('int64'), 'fastwam_replan_steps': Value('int64'), 'reset_order': List(Value('string')), 'seed': Value('int64'), 'startup_wait': {'counts_toward_policy_step_budget': Value('bool'), 'dummy_action_env': List(Value('float64')), 'steps': Value('int64')}, 'step_budget_by_suite': {'libero_10': Value('int64'), 'libero_goal': Value('int64'), 'libero_object': Value('int64'), 'libero_spatial': Value('int64')}, 'task_environment_lifetime': Value('string'), 'task_order_index': Value('int64'), 'trial_order': Value('string')}, 'schema_version': Value('string'), 'source': {'libero_commit': Value('string'), 'relevant_paths_clean': Value('bool'), 'task_map': {'bytes': Value('int64'), 'path': Value('string'), 'sha256': Value('string')}}, 'status': Value('string'), 'tree_hashes': {'algorithm': Value('string'), 'bddl_tree_sha256': Value('string'), 'format': Value('string'), 'init_states_tree_sha256': Value('string'), 'payload_tree_sha256': 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.
Paired LIBERO RGB and Simulator GT
This release pairs both RGB views, robot state, action, and simulator-derived GT at the same retained official HDF5 state. It preserves FastWAM's complete four-suite training membership: 1,712 episodes and 277,713 frames.
Training population
| Suite | Episodes | Frames | Tasks |
|---|---|---|---|
| libero_spatial | 434 | 53,229 | 10 |
| libero_object | 457 | 67,309 | 10 |
| libero_goal | 433 | 52,895 | 10 |
| libero_10 | 388 | 104,280 | 10 |
| Total | 1,712 | 277,713 | 40 |
Evaluation population
eval/ freezes 40 ordered
LIBERO tasks and 50 official initial states per
task: 2,000 online policy rollouts. It does not
contain prerecorded trajectories. The protocol uses seed 42, 30 startup wait steps,
a 700-step budget for libero_10, and 400 steps for the other suites.
The eval source is pinned to LIBERO commit 8f1084e3132a39270c3a13ebe37270a43ece2a01.
Layout and integrity
Each *_paired_gt_lerobot/ directory is a finalized v3 two-camera LeRobot suite.
Its meta/dataset_manifest.json, v3 episode status/input bindings, and v5 run
plan bind every training artifact and frozen source identity. The eval bundle
has its own manifest and per-file hashes.
All four suites are bound to EGL device 0 of
5 devices: NVIDIA Corporation / NVIDIA GeForce RTX 3070/PCIe/SSE2 /
4.6.0 NVIDIA 580.173.02.
Simulator GT uses the frozen legacy 8x8 bank followed by the dense 16x16 bank
(320 agent-view queries), plus the 16-offset post-noop FastWAM/LeRobot
dataset-step horizon bank. Its format is future-open-d4rt-paired-libero-gt-multigrid320-horizonbank16-v3. Apply its validity
masks during supervision.
Release payload tree SHA-256: a21212e56bd30e83538084e51d1a31dcb616f2e31793d9c1abd3b1b86e6a3742
The release seal embeds four full-decode PASS reports. Each report is bound
to its suite tree hash; the validation envelope also binds the final train,
eval, and complete payload tree hashes.
Validate all counts, paths, schemas, and hashes before use or upload:
python -m future_open_d4rt.tools.build_paired_libero_release \
--release-root . --validate-only
The GT validity masks must be applied during supervision. The eval seed bundle must be consumed as online rollouts; it is not a train/eval frame dataset split.
- Downloads last month
- 87