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
episode: int64
seg: int64
success: bool
steps: int64
pick_start_index: int64
transitions: int64
grasp_dist_m: double
lift_m: double
grasped: bool
target: string
demo_steps: int64
track_err_m_p50: double
track_err_m_max: double
wall_s: double
open_loop: bool
physics_hz: int64
grasp_step: null
finger_open_before_grasp_m: null
finger_open_end_m: double
obj_push_xy_before_grasp_m: double
obj_rise_before_grasp_m: double
action_saturated_steps: int64
load_s: double
frames: list<item: int64>
child 0, item: int64
end: int64
obj_grasp_m: double
eef_grasp_m: list<item: double>
child 0, item: double
eef_obj_m: list<item: double>
child 0, item: double
key_index: int64
reset_index: int64
start: int64
to
{'episode': Value('int64'), 'seg': Value('int64'), 'start': Value('int64'), 'end': Value('int64'), 'target': Value('string'), 'frames': List(Value('int64')), 'key_index': Value('int64'), 'reset_index': Value('int64'), 'eef_grasp_m': List(Value('float64')), 'eef_obj_m': List(Value('float64')), 'obj_grasp_m': Value('float64')}
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 544, 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 3052, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, 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 558, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, 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 361, 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 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
episode: int64
seg: int64
success: bool
steps: int64
pick_start_index: int64
transitions: int64
grasp_dist_m: double
lift_m: double
grasped: bool
target: string
demo_steps: int64
track_err_m_p50: double
track_err_m_max: double
wall_s: double
open_loop: bool
physics_hz: int64
grasp_step: null
finger_open_before_grasp_m: null
finger_open_end_m: double
obj_push_xy_before_grasp_m: double
obj_rise_before_grasp_m: double
action_saturated_steps: int64
load_s: double
frames: list<item: int64>
child 0, item: int64
end: int64
obj_grasp_m: double
eef_grasp_m: list<item: double>
child 0, item: double
eef_obj_m: list<item: double>
child 0, item: double
key_index: int64
reset_index: int64
start: int64
to
{'episode': Value('int64'), 'seg': Value('int64'), 'start': Value('int64'), 'end': Value('int64'), 'target': Value('string'), 'frames': List(Value('int64')), 'key_index': Value('int64'), 'reset_index': Value('int64'), 'eef_grasp_m': List(Value('float64')), 'eef_obj_m': List(Value('float64')), 'obj_grasp_m': Value('float64')}
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.
R1Pro frying-pan pick, precise phase (BEHAVIOR-1K 2026 challenge, cook_broccolini)
Demonstrations of the last, precise part of a right-hand frying-pan pick, converted from the BEHAVIOR-1K 2026
challenge R1Pro demos (behavior-1k/2026-challenge-demos annotations, behavior-1k/2026-challenge-rawdata
simulator states; MIT, see LICENSE) for policy learning (HIL-SERL / pi0.5 EXPO-FT) in
github.com/kwk2696/agentic-behavior-1k, branch hil-serl (hilserl/demos/convert.py, commit b6489b4 with the
frying-pan selection relaxed to every segment that lifts >= 5 cm).
- Task
cook_broccolini(task index 96), skillpick_up_from, objectfrying_pan, right hand. - 87 segments selected (all right-hand pan picks of the 200 episodes whose demo lifts the pan >= 5 cm), 66 kept, 3,121 transitions (per demo: median 40 steps, 10 Hz).
How a demo was made
- Pick start: the first 10 Hz frame at which the right hand's EEF is within 0.10 m of the grasp point (the EEF pose at the first gripper-close command, held in the pan's frame).
- Each segment was replayed in OmniGibson 3.9.3 / Isaac Sim 5.1 (physics 120 Hz, control 10 Hz) from its segment start: the whole robot up to the pick start, then base, trunk and left arm held still and only the right arm (IK pose delta) and right gripper driven (arm targets moved by the demo's base motion).
- A demo is kept only if that replay succeeds: the right hand grasps the pan and lifts it >= 5 cm. Only the steps from the pick start on are stored. Observations are re-rendered in the simulator.
demos.pkl
A Python list of transitions (pickle; hil-serl demo-buffer format, hilserl/common/transitions.py):
| key | content |
|---|---|
observations, next_observations |
head: (1, 128, 128, 3) uint8, right_wrist: (1, 128, 128, 3) uint8, state: (1, 25) float32 |
actions |
(7,) float32 in [-1, 1]: right-arm EEF delta in the robot base frame [dx, dy, dz] / 0.10 m, axis-angle [rx, ry, rz] / 0.25 rad; right gripper -1 close / +1 open on the step its command changes, else 0 |
rewards, masks, dones |
1 / 0 / True on the last (success) step, else 0 / 1 / False |
infos |
episode (demo_<raw episode>_s<segment>), step, success-predicate values, is_intervention (True) |
grasp_penalty |
-0.02 on a gripper command change, else 0 |
state (robot proprioception, robot base frame): eef_left_pos 3, eef_left_quat 4 (xyzw), eef_right_pos 3,
eef_right_quat 4, gripper_left_qpos 2, gripper_right_qpos 2, trunk_qpos 4, base_qvel 3.
Other files
selection.json: the 87 segments (raw episode id, segment index, start/end frame at 30 Hz).results/ep<E>_s<K>.json: each segment's replay result (success, lift, grasp, tracking error, pick start index).distances/dist_ep<E>_s<K>.json: per-10-Hz-frame distance of the EEF to the grasp point and to the pan.
- Downloads last month
- 5