Dataset Viewer
Duplicate
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:    ValueError
Message:      Illegal slicing argument for scalar dataspace
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 83, in _generate_tables
                  pa_table = _recursive_load_arrays(h5, self.info.features, start, end)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 267, in _recursive_load_arrays
                  arr = _recursive_load_arrays(obj, features[path], start, end)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 267, in _recursive_load_arrays
                  arr = _recursive_load_arrays(obj, features[path], start, end)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 267, in _recursive_load_arrays
                  arr = _recursive_load_arrays(obj, features[path], start, end)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 269, in _recursive_load_arrays
                  arr = _load_array(obj, path, start, end)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 236, in _load_array
                  arr = dset[start:end]
                        ~~~~^^^^^^^^^^^
                File "h5py/_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
                File "h5py/_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
                File "/usr/local/lib/python3.14/site-packages/h5py/_hl/dataset.py", line 931, in __getitem__
                  selection = sel2.select_read(fspace, args)
                File "/usr/local/lib/python3.14/site-packages/h5py/_hl/selections2.py", line 101, in select_read
                  return ScalarReadSelection(fspace, args)
                File "/usr/local/lib/python3.14/site-packages/h5py/_hl/selections2.py", line 86, in __init__
                  raise ValueError("Illegal slicing argument for scalar dataspace")
              ValueError: Illegal slicing argument for scalar dataspace

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.

Geodesic

Cosmic — MolmoSpaces Benchmark Evaluation Artifacts

Complete evaluation artifacts for Cosmic, the TAMP + foundation-model manipulation policy developed by Geodesic, evaluated on the MolmoSpaces benchmark suite (Franka arm, joint-position action space). Evaluations ran 2026-09-16 → 09-18 on eight RTX 5090 instances. Policy code is not open-sourced.

Results (oracle success condition, official eval_to_csv.py)

Benchmark Episodes Success rate
Pick-v1.1 (bench-v1) 1000 62.6%
Pick-v1.5 (bench-v2) 1000 61.1%
Pick-v2-classic 1000 51.4% (514/1000; one episode unrunnable, see below)
Pick-v2-filament 1000 49.1%
Pick-v2-rand-cam 1000 53.0%
Pick&Place-v1 1000 44.7%
Pick&Place-v2 1000 40.7%
Pick&Place-color-v2 1000 41.7%
Pick&Place-NextTo-v2 1000 43.3%
Open-v1 1000 58.5%
Close-v1 915 87.9%
11-task average 54.0%

Result CSVs are also published at https://github.com/Vrushabh27/cosmic_results.

Layout

  • Cosmic_<bench>_oracle.csv — leaderboard CSVs (also under <bench>/), Cosmic_all_csvs.zip — all 11.
  • <bench>/manifest.json — the scored episode set: one entry per benchmark episode with the h5 file + trajectory group that was scored and its success flag.
  • eval_output/<pod>/cosmic_<bench>/…/house_<i>/*.h5 — every trajectory produced by the campaign (raw harness output, ~10.9k scored episodes plus superseded duplicate runs). No videos were kept for this campaign.
  • logs/ — fleet guardian log, hourly coverage tables, sample lane logs; final_score.json — scoring summary.

Scoring notes

  • Exactly one instance per benchmark episode; an episode is identified by its house and object start pose. Where an episode was run more than once (chunk requeued after an infrastructure failure, or a two-episode house re-run to recover its sibling), the earliest completed run is scored — never the best of several.
  • Pick/PnP/NextTo CSVs come from the official scripts/benchmarks/eval_to_csv.py --success-condition oracle on the deduplicated trajectory set; Open/Close use a per-episode matcher against the DataGenConfig benchmark definitions with the same success condition and confidence interval.
  • Pick-v2-classic episode 394 (house 2695) cannot be simulated with the published benchmark archive (objaverse asset af3281f986cc40b9b3cbca1f72e77f46_visual.obj missing) and is counted as a failure (514/1000); the CSV's OVERALL row lists the 999 simulated episodes.
  • ~780 pick-benchmark episodes early in the campaign ran while the GraspGen-X grasp server was down (M2T2 only); they scored 2–4 points below the remainder (within noise) and were kept as-is.
Downloads last month
35