The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: ValueError
Message: Dataset 'workspace_bbox' has length 6 but expected 4
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/hdf5/hdf5.py", line 76, in _generate_tables
num_rows = _check_dataset_lengths(h5, self.info.features)
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 353, in _check_dataset_lengths
raise ValueError(f"Dataset '{path}' has length {dset.shape[0]} but expected {num_rows}")
ValueError: Dataset 'workspace_bbox' has length 6 but expected 4Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
rlbench_da3_offline_n50
Offline novel-view augmentation cache used to train the VM-VLA RLBench checkpoints
main_cvtgt-fixed-4gpu_50000 and baseline_5_posecls_b32_20000 in
mickeykang/openvla-rlbench-ablations.
For every (episode, frame) of the rlbench_mv train split, 50 novel views were rendered with
Depth-Anything-3 (depth-anything/DA3NESTED-GIANT-LARGE-1.1) from the 4 input cameras
(cam_mid_1, cam_mid_3, cam_original, cam_top_down) with camera perturbations of
radius ±0.20 m, azimuth ±20°, elevation ±20°, rendered at 224×224.
Layout
config.json
episode_0000.hdf5 ... episode_0999.hdf5 # 1000 episodes, 37.6 GB total
Generated with vla-scripts/generate_offline_da3_rlbench.py in the
MemVLA-Adapter repo
(--num_novel 50 --dr 0.20 --daz 20 --del 20).
Usage
hf download --repo-type dataset mickeykang/rlbench_da3_offline_n50 \
--local-dir $HOME/ws/data_rlds/rlbench_da3_offline/train_v2_dr20_az20_el20_n50
Then pass the directory as --da3_offline_hdf5_dir to vla-scripts/finetune.py in
mickeykang/openvla_rlbench (see REPRODUCE.md §13).
The cache is aligned per (episode, frame) with the RLDS dataset
mickeykang/rlbench_mv loaded with shuffle=False;
it is only meaningful together with that dataset.
Companion caches (eval split, and the 5-/10-view train caches used by the other baselines) are in
mickeykang/rlbench_da3_offline_aux.
- Downloads last month
- 214