The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: ValueError
Message: Dataset 'action57' has length 140 but expected 141
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 'action57' has length 140 but expected 141Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
tabletopsim-watch
Tabletop-Sim ALOHA demonstrations converted to the EgoVerse 57-d state/action format, with the retrieval manifest used for one-shot conditioning. This is the training-ready form: the raw simulator HDF5s (~112 GB) are not needed.
| count | |
|---|---|
| episodes | 260 |
| windows | 52,043 |
| queries | 52,043, top-6 matches each |
| tasks | box_into_pot_easy 60, dish_drainer 50, handover_box 50, lift_box 50, shoes_table 50 |
Six box_into_pot_easy colour combinations (yellow/white/brown box × blue/green pot) are separate
language instructions, 10 episodes each; the other four tasks have one instruction apiece.
Layout
data/<task>/episode_<i>.h5 # 260 episodes, EgoVerse-57 layout
manifest/
windows.npz # ep / w0 / w1 / emb / cls per window ("ep" = "<task>/episode_<i>")
pairs.npz # query_idx, match_idx (N, 6), match_dist
state57_stats_pctl.json, abs_action57_stats_pctl.json, action57_stats.json
state20_stats_pctl.json, abs_action20_stats_pctl.json # 20-d variant
t5_embeddings.pkl # per-instruction T5 embeddings
tasks.json # task -> instruction text
stats.json # the table above, machine-readable
Use
Point the training config at the unpacked root and the manifest:
TABLETOP_ROOT = <...>/data
manifest_dir = <...>/manifest
windows.npz:ep is "<task>/episode_<i>", matching data/<task>/episode_<i>.h5 with no renaming.
Both the 57-d and the 20-d normalization statistics are included; pick the pair that matches the
target_group your run uses (abs_action57 for the absolute-command recipe).
Provenance
Converted from the Tabletop-Sim ALOHA HDF5 release with convert_tabletop_to_egoverse57.py
(260 episodes → 52,043 windows), then T5 instruction embeddings were built from tasks.json.
- Downloads last month
- 20