rubatotree/act-pick-black-cat
Robotics • 51.7M • Updated • 13
Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception: SplitsNotFoundError
Message: The split names could not be parsed from the dataset config.
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 290, in _generate_tables
pa_table = paj.read_json(
io.BytesIO(batch), read_options=paj.ReadOptions(block_size=block_size)
)
File "pyarrow/_json.pyx", line 342, in pyarrow._json.read_json
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
return check_status(status)
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowInvalid: JSON parse error: Column() changed from object to number in row 0
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
for split_generator in builder._split_generators(
~~~~~~~~~~~~~~~~~~~~~~~~~^
StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 101, in _split_generators
pa_table = next(iter(self._generate_tables(**splits[0].gen_kwargs, allow_full_read=False)))[1]
~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 304, in _generate_tables
batch = json_encode_fields_in_json_lines(original_batch, json_field_paths)
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 111, in json_encode_fields_in_json_lines
examples = [ujson_loads(line) for line in original_batch.splitlines()]
~~~~~~~~~~~^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 20, in ujson_loads
return pd.io.json.ujson_loads(*args, **kwargs)
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
ValueError: Expected object or value
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 68, in compute_split_names_from_streaming_response
for split in get_dataset_split_names(
~~~~~~~~~~~~~~~~~~~~~~~^
path=dataset,
^^^^^^^^^^^^^
config_name=config,
^^^^^^^^^^^^^^^^^^^
token=hf_token,
^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
info = get_dataset_config_info(
path,
...<6 lines>...
**config_kwargs,
)
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.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.
Pick black cat into the green SIGGRAPH mug
1280 synthetic 12-second demonstrations, 25 Hz, with synchronized 640×480 front wrist and side RGB cameras and six SO101 joints. The target is the small black plush cat with yellow eyes; the larger black cat with glasses and the glass cup are distractors. Success includes resting inside the green mug or stably across its rim.
The scene was reconstructed from multiview photographs. Motion derives from documented physical seed simulation and checked kinematic retargeting. Most derivatives received lightweight lineage, kinematic and native-data checks, rather than a new physics simulation or full manual RGB review. Per-episode provenance records the exact scope.
The fixed dataset contains 64 Cycles and 1216 Eevee episodes, with 256 weak and 1024 strong augmentations. Eevee sample counts vary and are explicitly labeled. Camera ranges changed during production: later weak placement augmentation samples 25–40 cm horizontal translation, absolute yaw changes of 12.5–20 degrees, absolute pitch changes of 5–10 degrees, and roll within ±2.5 degrees, moving the camera support together. Earlier prepared episodes retain their earlier ranges. The eight-episode limit applies to the special opposite/overhead camera templates, not to every large displacement. Per-episode specifications record the actual parameters and revision. Mixed augmentation includes motion, tabletop object placement, camera pose, materials and lighting. Some long-tail observations intentionally contain severe exposure or appearance changes.
Data contract
LeRobot v3 format; action has six absolute joint motor reference angles in radians, targeting the next 25 Hz endpoint. It is distinct from measured joint state. The original controller interpolates motor references at 500 Hz. Do not directly interpret these values as degrees or motor percentages. labels.json and per-episode provenance document quality, nuisance strength and inherited state.
Use observation.images.front, observation.images.side and observation.state as policy inputs. Simulation annotations and hidden state are metadata, not deployment observations. The initial search segment precedes target-dependent reaching.
Validation and limits
The native aggregate was read with official LeRobot, all numeric rows and episode action boundaries were checked, and all videos fully decoded. aggregate_validation.json contains results. All derivatives share one source episode group; the train metadata does not establish an independent test split. Rendered-data validation and offline fitting errors do not establish real-robot task success or policy generalization. No policy rollout success rate is available at dataset release.
Reproduction
provenance/episode_map.json maps global indices to source shards, randomization and seed lineage. The released provenance includes generation parameters, relevant code snapshots and validation scope. The public dataset ID is rubatotree/miracle-pick-black-cat.
- Downloads last month
- 62
Models trained or fine-tuned on rubatotree/miracle-pick-black-cat